Rocket Science: Chapter 1: A Toxic Legacy

Sources of drinking water for almost 7 million Californians and unknown millions of other Americans are contaminated with perchlorate, a chemical that disrupts normal thyroid function, may cause thyroid cancer and persists indefinitely in the environment, but is unregulated by the state or federal government.

An Environmental Working Group investigation found that the state has known about contamination of California groundwater by perchlorate, the main ingredient in rocket and missile fuel, for almost 50 years. Groundwater is an important source of drinking water in California, contributing about one-sixth of the state's drinking water during an average year, and almost one-third in drought years. (CADWR 1998.) Yet today, state and federal regulators are still dragging their feet on setting safety standards for perchlorate in drinking water, in part because of concerted pressure from the U.S. military, the space program and major defense contractors. As a result, standards that emerge are unlikely to provide adequate protection for developing fetuses, infants or children.

The human thyroid gland controls growth, development and metabolism. Perchlorate affects the thyroid because it is taken up preferentially by the gland in place of iodide, a necessary nutrient. This, in turn, can affect thyroid hormone levels. An underactive thyroid gland in adults can lead to fatigue, depression, anxiety, unexplained weight gain, hair loss, and low libido. More serious, however, are the effects of thyroid hormone disruption in the developing fetus and child. Small changes in maternal thyroid hormone levels during pregnancy have been associated with reduced IQs and attention deficit in children. Fetuses, infants and children who experience bigger changes in hormone levels may suffer mental retardation, loss of hearing and speech, or deficits in motor skills.

There are many scientific unknowns surrounding the health effects of perchlorate. To date, none of the scientific studies performed on the effects of perchlorate have adequately addressed how perchlorate might affect neurological development of children whose mothers were exposed to perchlorate while pregnant. Nor has there been any research conducted to determine whether perchlorate is concentrated in breast milk, which is considered a distinct possibility, if not likely. (EPA 1999c.)

What has been documented through animal research, however, is that perchlorate can disrupt the thyroid hormone system at low levels of exposure, and that this may also lead to thyroid cancer, the disruption of female menstrual cycles, and the weakening of the immune system at higher levels. (EPA 1998.)

As of June 2001, perchlorate has been detected in 197 water sources in 58 different public water systems in California, in some cases at levels far above the state considers safe to drink. These systems serve almost 7 million people, mostly in the San Gabriel Valley and Inland Empire regions of Southern California and the Rancho Cordova area of Sacramento County. (Figure 1, Table 1.) In addition, perchlorate was detected in 56 other water sources that have only been tested once, so the state does not consider the contamination to be confirmed. The state has so far sampled only about 15 percent of California's drinking water sources, raising the strong likelihood of undetected contamination (CADHS 2001).

Significant concentrations of perchlorate are also found in the Colorado River and in Lake Mead. The lower Colorado supplies drinking water to more than half of Southern California, and both the river and Lake Mead are major sources of drinking water for Arizona and Nevada, including Phoenix and Las Vegas. The Colorado also irrigates almost one million acres of some of the nation's most productive farmland in California and Arizona, raising concerns about the toxicity of lettuce and other crops that may absorb perchlorate.

Nationwide, the U.S. Environmental Protection Agency (EPA) has found perchlorate-tainted water in 18 states and believes it poses a threat to drinking water wherever rocket fuel or rockets were made, tested, or disposed, and wherever other large scale uses of perchlorate are found - as many as 39 states in all. (Figure 2, Table 2.) So far, only a small percentage of these sites have been tested for contamination. According to the EPA, however, soil and/or groundwater contamination have been found at essentially all such locations where an effort to test has been made. (EPA 2001a.) According to the EPA, "[A]t essentially every listed facility where an effort has been made to test for perchlorate, perchlorate has been found in the soil or groundwater."

In 1999 the EPA listed perchlorate under the federal Unregulated Contaminant Monitoring Rule, with monitoring beginning in January of 2001. As a result, all large public water systems and a sample of small systems nationwide are required to conduct a one-year, one-time monitoring program for perchlorate before the end of 2003.

Perchlorate contamination of California's water was first detected almost 50 years ago, and perchlorate's damaging effect on the thyroid has been known for just as long. But neither the California Department of Health Services (DHS) nor the EPA has established any enforceable health standard for safe levels of perchlorate in drinking water.

The state has set a so-called action level for perchlorate in drinking water, above which it recommends that water suppliers shut down the contaminated source. But "action level" is a misnomer. Although many drinking water sources where perchlorate has been detected above the action level have been voluntarily closed, there is no enforceable requirement that suppliers close the source, or even that they inform people who have been drinking the water. The EPA was expected to issue its latest "provisional" national standard for perchlorate in drinking water this year but delays in the submittal of data have "postpone[d] the assessment indefinitely." (EPA 2001a.) It is predicted to be several more years before there is an official state or federal drinking water standard.

It is already clear that the EPA's new provisional standard not only will lack teeth, but will be set at a level that will not protect the public, particularly children, from harmful levels of perchlorate in drinking water. EWG calculated that the EPA's most likely proposed standard would leave infants who are fed formula mixed with tap water - about 40 percent of all babies - exposed to between 7.5 times to more than 2000 times the safe level of perchlorate, depending on the margin of safety applied. EWG's calculations show that current levels of perchlorate found in Lake Mead and the Colorado River are not safe for drinking water.

EWG analysis shows that bottle-fed infants who consume water in formula contaminated with perchlorate at the EPA's current proposed level (32 ppb) would be getting a dose of perchlorate only nominally lower than may affect thyroid hormone levels (40 ppb). Breastfed infants are also at risk because small changes in maternal thyroid hormones during pregnancy can have adverse effects on the fetus and because perchlorate may be passed on or concentrated in breast milk. (Escobar et al. 2000; EPA 1999c)

Each time the federal government has set a new provisional perchlorate safety standard, the recommended "safe" levels have been raised. The EPA's first provisional standard, set in 1992, was 4 parts per billion (ppb) of perchlorate in drinking water. In 1995, the provisional standard was revised as a "safe" range of 4 ppb to 18 ppb. In 1998, the EPA's provisional standard was raised to 32 ppb. The California action level, set in 1997, is 18 ppb, and the level at which the state recommends, but does not require, shutting off the contaminated source of water is 40 ppb. (Figure 3.)

But in calculating the previous provisional standards, which the state relied on in setting its action level, the EPA ignored critical data, neglected the needs of sensitive populations, and significantly underestimated the scientific uncertainties surrounding perchlorate's health effects.

The EPA ignored a study that found perchlorate to significantly affect thyroid hormones at a level ten times lower than the study on which it based the 1998 provisional standard. Most significantly, the agency based the provisional standards on adult rather than infant or child body weights, despite recognizing that perchlorate "may pose a serious threat to developmental processes in children." (EPA 2001a.)

While regulators have been steadily relaxing standards, medical research has been continually lowering the threshold at which perchlorate is known to have harmful effects.

Last year, a study by researchers with the Arizona state health department, of newborn infants whose mothers drank water from Lake Mead, found abnormal levels of a thyroid hormone in babies whose mothers were exposed to levels of perchlorate that were less than one-fifth the current EPA provisional standard and only one-third of the California action level. EWG calculates that to adequately protect infants and children, perchlorate standards should be at least four times, and as much as 1,000 times, more stringent than California's current action level. (Table 3.)

Many sites of perchlorate pollution are contaminated far above levels that are known to have negative health effects. In California, concentrations of perchlorate in groundwater at Superfund cleanup sites have been detected at hundreds of thousands of parts per billion. According to the most recent data from the Department of Health Services, the average concentration in contaminated sources is 83 ppb in Sacramento County, 36 ppb in San Bernardino County and 14 ppb in Los Angeles County.

Some contaminated wells have been shut down, but many are still in service. In Sacramento County, three of the 15 known contaminated sources are still delivering tap water to homes. In San Bernardino County, 20 of 37 contaminated wells are active, and in Los Angeles County, 77 of 99 contaminated sources are active. Statewide, 70 percent of known contaminated sources are still delivering water to customers. And sources that are listed as "inactive" may not be permanently closed, just not used for 12 consecutive months. (Table 3.)

In November 2000, concentrations of perchlorate in already treated drinking water drawn from Lake Mead reached a high of 24 ppb, prompting the Las Vegas water utility to temporarily shut off intake from the lake. Since monitoring of Lake Mead began in 1997, perchlorate levels in treated water from the lake have averaged 9 ppb, about the same level currently found in the lower Colorado River. Perchlorate levels as high as 1,700 ppb have been detected in the Lake Mead drainage basin.

Concerted pressure to set a looser, less protective drinking water standard is being applied at both the state and federal regulatory levels by a powerful alliance of opponents: the sole remaining U.S. manufacturer of perchlorate; giant aerospace contractors, whose rocket-making and -testing sites, and the communities surrounding them, are contaminated with the chemical; and the U.S. Air Force, which commissioned the contractors' work with perchlorate.

In California the defense contractors responsible for most perchlorate contamination are Lockheed Martin of Bethesda, Md., in the San Gabriel Valley and Inland Empire, and Aerojet, a division of Sacramento-based GenCorp, in Sacramento County and the San Gabriel Valley. They are also major players in the Perchlorate Study Group, a chemical industry task force that, in conjunction with the Air Force, sponsors perchlorate research and meetings with regulators for the explicit purpose of delaying or weakening proposed standards.

In a scientifically invalid and morally unethical attempt to prove that perchlorate isn't so dangerous, Lockheed Martin, other companies in the Perchlorate Study Group and the Air Force have sponsored tests in Southern California, Oregon and other locations in which human subjects were paid to ingest daily doses of perchlorate. In one study, which began in August 2000 at Loma Linda University Medical School near San Bernardino, so-called "volunteers" were paid $1,000 to ingest, every day for six months, doses of perchlorate up to 83 times higher than the safe daily intake recommended by the State of California.

In November 2000, EWG obtained and published Loma Linda documents that showed these human guinea pigs were not fully informed of the dangers of perchlorate or the tests' true purposes: To help the Perchlorate Study Group resist stricter regulations and help Lockheed fight lawsuits from San Bernardino County residents whose drinking water was contaminated by perchlorate. (EWG 2000.) A month later, the U.S. Office for Human Research Protections launched an investigation, still ongoing, to determine whether the Loma Linda tests violate the ethical standards required for research facilities with federal contracts.

Perchlorate contamination is not removed by conventional water treatment processes, nor by technologies that remove other common groundwater pollutants. Cleaning up the perchlorate mess, therefore, is expected to be lengthy and expensive. Groundwater cleanup at just one part of the Aerojet Superfund site in Sacramento County is predicted to cost at least $55 million and, incredibly, take 240 years to complete. If a less protective state or federal drinking water standard is set, Aerojet and the parties responsible for perchlorate contamination will save millions of dollars in cleanup costs.

But no matter how strict or lax the eventual standards are, the defense contractors - who for decades have pocketed big profits from their government-funded perchlorate work - have cut deals with the Defense Department that will stick taxpayers with most of the costs of cleanup. Two years ago, Aerojet negotiated an extraordinarily favorable deal with the government in which 88 percent of cleanup costs are recoverable from future military contracts. (Aerojet 2000.) The exact percentage negotiated by Lockheed Martin is secret, but knowledgeable sources say Lockheed's deal is equal to or better than Aerojet's. (EWG 2001.) The eventual cost to U.S. taxpayers of cleaning up the mistakes of the military and its contractors will reach billions of dollars.

Recommendations

• The State of California and the EPA should act promptly to set and enforce health-based safety standards for perchlorate in drinking water.
• State and federal standards for perchlorate in drinking water should be fully protective of public health, especially the health of developing fetuses, infants and children. The standard should also take into consideration all relevant scientific data, the needs of hypothyroid individuals, and the significant uncertainties. The drinking water standard should be no higher than 4.3 ppb.
• All human testing of perchlorate should be stopped immediately and tests on humans of any pollutant, pesticide or other toxic chemical should be banned.
• The Air Force, or any other branch of the U.S. military, should be prohibited from lobbying to weaken existing or proposed environmental laws or regulations.
• The corporations responsible for perchlorate contamination, and who have profited immensely from government contracts, should pay their share of cleanup. The U.S. military clearly shares responsibility, but making taxpayers pay for 90 percent of the cleanup of operations that made aerospace giants millions in profits is an egregious example of the "double-dipping" defense contractors are notorious for.
• Perchlorate polluters and the government should invest significantly more money into research on new cleanup technologies, and cleanup schedules must be greatly accelerated.