Weed Killers By The Glass
September 1, 1995

Weed Killers By The Glass: Chapter 3. Findings and Recommendations

The results of Weed Killers By The Glass substantially confirm the Environmental Working Group's 1994 findings of Tap Water Blues (Wiles et al. 1994). Millions of individuals in the Midwestern United States are commonly exposed to one or more pesticides in a single glass of tap water. During peak runoff periods pesticide contamination levels repeatedly exceed federal health standards and pose significant health risks.

Some water utilities, notably the Kansas City, MO water system, have taken a public position that they will not allow atrazine contamination to exceed EPA lifetime standards for even one day. And the American Water Works Association (AWWA), in their comments to the EPA on the atrazine special review, noted that "AWWA is concerned with exceedances of any MCL at any time..." (AWWA 1995; emphasis in original). Nonetheless, our testing of tap water in 29 cities revealed:


1. Weed Killers are Routinely Found in Tap Water at Levels that Exceed Federal Health Standards.

It is not unusual for atrazine and cyanazine levels to exceed EPA Maximum Contaminant Levels (MCL), or Lifetime Health Advisories (LHA) in treated tap water. Cyanazine levels exceeded federal health guidelines in 35 percent of all samples. Atrazine concentrations were above the federal health standards in 17 percent of all samples.

In some cities, herbicides in tap water exceed federal health standards for weeks or months at a time. In Springfield, IL, cyanazine exceeded the LHA in every sample that was collected. In eight other cities -- Danville and Decatur, IL; Indianapolis, IN; Columbus and Bowling Green, OH; Jefferson City MO; New Orleans, LA; and Kansas City, KS -- more than half of the samples that were collected exceeded the atrazine MCL or the cyanazine LHA.

The cyanazine LHA was exceeded at least once in 18 cities (62 percent) with a total population of 4.48 million people, 60 percent of the population covered in the study (Table 5). Atrazine exceeded the MCL at least once in 13 communities with a total population of 2.98 million people, representing 43% of the cities, and 40% of the of the population for the study area (Table 6).

Danville, Illinois has the worst cyanazine problem of the cities studied. Six tap water samples from Danville contained cyanazine at more than 10 times the LHA and in one sample cyanazine was found at 34 times the LHA. The nine samples with the highest cyanazine concentrations all came from Danville, followed by tap water samples from Kansas City, KS; Decatur, IL; Fort Wayne, IN; and Omaha, NE.

The highest atrazine concentration -- 18 ppb more than 6 times the MCL -- was also found in Danville, IL. In fact, the six samples with the highest atrazine concentrations all came from Danville. The four communities with the next four highest atrazine detections, all above the MCL, were Fort Wayne, IN; Columbus and Bowling Green, OH; and Kansas City, KS.


2. Contamination is ubiquitous.

Atrazine was found in tap water in 28 out of 29 cities tested, cyanazine was found in 25, metolachlor in 19, acetochlor in 15, alachlor in 10, simazine in four and metribuzin in two. One of two breakdown products (known as metabolites) of atrazine, desethylatrazine and desisopropylatrazine, was found in 12 communities (Table 7). The only community where weed killers were not found in tap water was Memphis, TN, which obtains its tap water from deep groundwater wells.


3. Average Levels of Weed Killers in Tap Water Exceeded Federal Standards in 13 Cities.

Average cyanazine levels in tap water exceeded the LHA in 13 cities during the entire testing period (Table 8). Average atrazine levels in drinking water during the same period exceeded the federal MCL in six of these same 13 cities, including Indianapolis IN; Columbus OH; and Fort Wayne IN (Table 9).

Combined levels of cyanazine and atrazine (total "triazine" level) exceeded the atrazine standard of 3 ppb in 11 cities tested. This is significant because the EPA has determined that the cancer risk from these herbicides is additive (PD1). Cyanazine, however, is a significantly more potent carcinogen than atrazine. Any combination of cyanazine and atrazine that exceeds the atrazine standard, therefore, is actually more toxic than exposure to atrazine alone at the same level.


4. Routine Exposure to Many Different Pesticides in a Single Glass of Water.

Tap water from two-thirds of the cities tested contained at least four and as many as nine pesticides or pesticide by-products (Table 10). Current EPA drinking water standards are set one chemical at a time and assume that the simultaneous exposure we document in this study does not occur.

Samples of tap water were tested once a month (twice during the study period) for eleven different pesticides. These tests found two or more pesticides or pesticide metabolites in the drinking water of 27 out of 29 cities, three or more in 24 cities, four or more in 21 cities, five or more in 18 cities, six or more in 13 cities, and seven or more pesticides or metabolites in the treated tap water of five cities -- Ft. Wayne, IN; Muncie, IN; Danville, IL; Columbus, OH; and Bowling Green, OH (Table 10). In Fort Wayne, IN, nine different pesticides and metabolites -- atrazine, cyanazine, metolachlor, alachlor, metribuzin, acetochlor, desethylatrazine, desisopropylatrazine, and simazine -- were found in a single sample of tap water collected in June, 1995. Three of these pesticides were found at levels above EPA standards. The nine pesticides included two probable human carcinogens, five possible human carcinogens, one pesticide responsible for birth defects, and four pesticides that disrupt the endocrine or hormone system.


5. Infants and Children are Exposed at Unsafe Levels and Mixtures of Pesticides in Infant Formula, Juices, and Drinks Reconstituted with Tap Water.

Drinking water standards do not account for the vulnerability of infants to toxic chemicals such as these weed killers, nor do they account for the high volume of water children drink relative to adults.

We estimate that 45,000 infants in 29 cities drank infant formula reconstituted with tap water contaminated with weed killers during this six week period. More than 10,000 infants drank infant formula made with tap water contaminated with atrazine at levels above the federal MCL. These same 10,000, plus an additional 8,400 infants, drank infant formula made with tap water contaminated with cyanazine at levels above the federal LHA. An estimated 28,000 infants drank infant formula reconstituted with tap water that was contaminated with at least four pesticides, and as many as nine pesticides and toxic pesticide by-products.




Federal Drinking Water Monitoring Requirements are Fundamentally Flawed.

Federal drinking water monitoring requirements provide regulators and public health officials with a fundamentally distorted picture of contamination levels in water. Extended periods of exposure above federal standards are not identified by federal monitoring requirements, nor are peak exposures that may exceed these standards by 10-fold to 30-fold or more.

Within the peak contamination period extended and repeated exposure to weed killers at levels above federal health standards is common in the cities we studied. Federal monitoring requirements, in contrast, treat all seasons the same and mandate only one sample during each quarter of the year, including the three month heavy contamination period. Even this lone sample can be taken before herbicides are applied and while contamination levels are low, or after pesticides have been largely flushed downstream.

On top of this, there is no monitoring requirement for so-called "unregulated contaminants" such as cyanazine, even though cyanazine was found at levels exceeding federal health advisories more often than any other herbicide.

These failings are particularly disturbing because federal drinking water standards do not incorporate safety factors to protect the public from extended periods of exposure above the MCL or LHA. The standards also do not consider the risks of exposure to multiple herbicides simultaneously, and they do not explicitly take into account the special risks to children. Finally, EPA's standard-setting methodology does not adequately protect the public from cancer risks.


Conventional Water Treatment Does Not Remove Weed Killers.

Following the release of Tap Water Blues, many water utilities claimed that the water treatment techniques they were using were able to adequately remove herbicides from contaminated source water. This is not the case.

All of the water tested in this study was treated tap water. In most cases, utilities are using only conventional water treatment -- chlorination and sand filtration -- which does nothing to reduce weed killer levels in water delivered to the community.

In a few communities, including Springfield, IL, and Kansas City, MO, water utilities are spending hundreds of thousands of taxpayer dollars using powdered activated carbon (PAC) in an attempt to reduce the levels of atrazine. While PAC can be used to reduce contamination levels of individual weed killers, it is not a solution to the problem. Springfield, Illinois, which utilizes vast amounts of PAC, still had the second highest cyanazine concentration of any city that was tested. PAC reacts differently with different chemicals, and its effectiveness in treating complex mixtures of contaminants in tap water is unproven.

In Bowling Green, the water utility uses a holding tank that they claim helps to reduce herbicide concentrations. In fact, the holding tank had just the opposite effect. By holding highly contaminated water in the tank, Bowling Green water authorities actually increased the levels of atrazine and cyanazine from below federal health standards on May 15, to levels three times higher than federal health standards by June 26th.

The only technology that can adequately remove pesticides once they have contaminated water supplies is the more expensive Granular Activated Carbon. Instead of costly technical fixes to the problem, the pollution prevention approach of restricting pesticide use is the most efficient and effective means of ensuring the safety of water supplies.




  • Parents in the most contaminated communities should seriously consider alternatives to tap water for infant formula, reconstituted juices or drinks for their infants and children from May 1 through August 30.

    The most contaminated cities identified in this study include:

    Danville, Decatur, Granite City and Springfield, Illinois
    Columbus and Bowling Green, Ohio
    Indianapolis and Fort Wayne, Indiana
    Kansas City, Kansas
    Jefferson City, Missouri
    Omaha, Nebraska
    New Orleans, Louisiana

    All of these cities had average contamination levels that exceeded at least one federal health standard for the study period, with the exception of New Orleans, LA, which did not exceed any individual health standard for the period but had a combined triazine average contamination level above the atrazine MCL of 3 parts per billion.


  • The EPA should require daily monitoring for triazine herbicides with inexpensive immunoassay tests for all surface-water-supplied drinking water systems in the corn belt. The monitoring cost is about $1,500 per city; less than 10 cents per person in a city of 20,000.


  • The EPA should phase out the use of the triazine herbicides by September 1996.
  • Congress must strengthen federal pesticide and drinking water laws so that they explicitly protect infants and children from acute and chronic effects of these contaminants.


  • Absent Congressional action, the EPA should move to set pesticide and drinking water standards to protect infants and children.


  • When setting drinking water standards to protect infants and children the EPA must strictly follow the recommendations of the National Academy of Sciences Report, Pesticides in the Diets of Infants and Children. At a minimum the EPA must specifically account for contamination of tap water with many different pesticides and metabolites. The agency (1) must explicitly account for additive or synergistic risks that may result from pesticides that act via a similar toxic mechanism or cause a similar toxic effect, (2) it must specifically account for any increased sensitivity or risk associated with infant or childhood exposure to these mixtures of compounds, and (3) it must consider all routes of exposure to the pesticides that a infant or child might encounter.