No Filter: Treating Drinking Water at the Faucet Is Rarely a Utility’s Contamination Solution

Under-the-sink filters are effective. But regulators impose stringent requirements on utilities that wish to use them.

A sign at a Rally's fast food restaurant in Flint, Michigan advertises filtered water. A sensible water quality solution for individuals or businesses, water filters are a challenge for utilities to deploy. Photo © J. Carl Ganter / Circle of Blue

A sign at a Rally’s fast food restaurant in Flint, Michigan advertises filtered water. A sensible water quality solution for individuals or businesses, water filters are a challenge for utilities to deploy. Photo © J. Carl Ganter / Circle of Blue

By Brett Walton, Circle of Blue

In response to the lead emergency in Flint, the state of Michigan distributed 146,843 water filters to residents. Installed under the sink, the devices effectively scrub lead from drinking water. U.S. Environmental Protection Agency testing in Flint shows lead concentrations in filtered water at roughly one-tenth of one part per billion — near complete removal. Marc Edwards, the Virginia Tech professor who brought national attention to Flint’s water failures, says that at roughly $US 40 per unit filters are an inexpensive solution and he recommends them “without reservations.”

And yet, Flint is an anomaly in this regard. Instead of making costly upgrades to central treatment plants, federal law allows utilities that serve fewer than 10,000 customers the option of filters to meet certain state and federal drinking water standards. But even though communities across the country are struggling with the cost of complying with drinking water regulation, few utilities use the devices. The EPA does not track national figures, but take Texas, for instance. Only one public water system — out of more than 4,600 — uses filters, according to the Texas Commission on Environmental Quality. And that system, an industrial facility, uses them on handwashing stations only.

Policy analysts and municipal officials say that even though the technology is proven, regulatory restrictions impose too daunting a barrier. Federal rules require utilities to own, operate, and maintain the filters, conditions that can require monthly home visits by utility employees or contractors to do water testing and check for broken parts. Labor costs and educating customers about the units are a drag on the utility budget, and some homeowners object to the intrusion into private space. In the end, high operating expenses mean that the numbers for filters often do not pencil out. Though a sensible option for individuals concerned about their tap water, filters rarely turn out to be the cheapest solution for a public water provider.

“The burden of ensuring their effectiveness and maintaining them make these devices an unattractive option,” William Diamond, a former U.S. Environmental Protection Agency drinking water division director, wrote in a 2001 memo.

Until recently little had changed. Then came the Flint scandal. The combination of high lead levels, poverty, and the lack of any other immediate fix has the EPA reevaluating, at least for lead, the use of filters to meet water quality goals.

Treatment Inside the Home

Filters, known as point-of-use (POU) treatment in regulatory jargon, were part of the 1996 amendments to the federal Safe Drinking Water Act. Those amendments addressed issues of concern for small communities, the biggest of which was the cost of meeting drinking water standards. The amendments allowed utilities the option of filters to cleanse drinking water of certain contaminants. Filters were not permitted for microbes, which cause immediate illness or risk of death — cryptosporidium or E. coli, for example — or for chemicals that are carried in water vapor and can be inhaled, such as radon.

States were free to set stricter standards and many did. California, for one, has a number of qualifications: the system must serve fewer than 200 connections and not have a feasible option for treating water at a municipal facility. In addition, the decision to install POU must not have “substantial” community opposition, defined as more than 50 percent of customers disapproving or not voting. POU permits extend for three years, after which the water system would have to file more paperwork to reapply.

The Coachella Valley Water District is one California jurisdiction that investigated POU treatment. California established new standards in 2014 for chromium-6, an industrial chemical, and roughly one-third of the district’s 98 wells exceeded the limit.

Though Coachella’s size — it serves more than 200,000 people in the Palm Springs area — disqualified it from the state’s POU requirements, district officials still wanted to assess the cost of faucet filters. If they turned out to be cheaper than central treatment, the officials might have lobbied for a change to the size restrictions.

Filters, in fact, were not cheaper. They were the least desirable of six options evaluated by Hazen and Sawyer, a New York engineering firm. The capital cost of installing POU – between $US 169 million and $US 363 million — was in line with the other options, which included a mixture of treating water at the wells or at a central plant. But the operating costs were outrageous: two to three times more expensive per year.

In the end, the report authors concluded that although filters were “technically feasible,” regulations inflated the operating cost beyond range.

Officials in Pretty Prairie, Kansas, also discovered the hurdles in place for filters as they sought in recent years to address nitrate contamination in the town’s well water.

“We researched it,” Patti Brace, the town clerk, told Circle of Blue. But the costs stacked up: $US 408,960 to purchase and install the units on Pretty Prairie’s 320 water meters, and at least $US 54,000 per year in operation and maintenance. Those figures were for filters on just one tap per home, meaning no protection for the bathroom faucets or showerheads. The town ultimately chose a reverse osmosis system, a $US 1.7 million option that was still cheaper in the long run.

“Filters were a high-maintenance solution,” Brace explained. “If we had only 25 to 30 homes, it would be different. But 300-plus homes was too far out of reach.”

Why is Flint different? For one, it was an emergency. Second, lead is in a different class than other contaminants. It is the only one regulated by the EPA that requires testing at the household tap. Because lead comes from service lines that direct water into the home, contamination happens after water exits the municipal treatment plant. For this reason, the Lead and Copper Rule, the federal standard, is called a “shared responsibility” standard. And that opens opportunities, Edwards says.

“Because of the shared responsibility structure, that means that you don’t have all these other requirements” — namely utility ownership and maintenance — “that have made filters less attractive in other cases,” Edwards told Circle of Blue.

The EPA is starting to think along those lines. In a policy paper published October 26, the agency noted that it is considering a role for faucet filters as a tool for addressing high household lead levels. Authorities could require a utility to provide filters to a household with lead service lines if construction is occurring nearby. (The pounding of backhoes and jackhammers can shake loose lead particles in the pipe.) Regulators could also order utilities to provide filters to homes that test high in lead, as a stopgap measure until the service line is replaced. The EPA expects to publish a draft rule by June.

Mike Keegan, an analyst with the National Rural Water Association, which represents small systems, recognizes that regulators are often wary of filters. But, he points out, if central treatment proves too costly, the alternative may be no treatment at all.

Edwards, for one, foresees a greater role for filters in addressing lead, noting that it will take utilities decades to replace the country’s more than 11 million kilometers (7 million miles) of lead service lines. “As an interim measure, it looks attractive,” he said.