The Power of Information

5 Years in a Row

Top rated, 2003-2007

Link to Charity Navigator

At EWG, our team of scientists, engineers, policy experts, lawyers and computer programmers pores over government data, legal documents, scientific studies and our own laboratory tests to expose threats to your health and the environment, and to find solutions. Our research brings to light unsettling facts that you have a right to know.

sign up

support ewg
printable all on one page email this page

Rocket Science: Chapter 3: Rocket Science


About 90 percent of the perchlorate manufactured each year goes to the Air Force, NASA and defense contractors to make rocket fuel. Ammonium perchlorate is the main component of solid rocket fuel, typically constituting 60 percent to 75 percent of missile propellant and about 70 percent of space shuttle rocket motors. (EPA 1998.) Perchlorate makes up the bulk of the 55 million pounds of fuel from decommissioned Cold War missiles that are currently awaiting disposal, an amount projected to increase to 164 million pounds by 2005. (Siddiqui et al. 1998.)

Perchlorate contamination of groundwater is the result of a combination of processes: manufacturing, rocket and fuel development, testing and maintenance. In each process, for more than 50 years the chemical industry, defense contractors or the military disposed of millions of pounds of perchlorate waste by simply flushing it with high-pressure water jets. The waste stream was enormous, because if a launch-ready rocket sits idle for too long, the fuel can go "flat" and hundreds of thousands of pounds of perchlorate must be replaced with a fresh supply. A space shuttle rocket motor, for instance, contains about 700,000 pounds of perchlorate. (NASA 1989.) Flushing generates large volumes of wastewater contaminated with perchlorate at levels up to 1 percent of the total volume. (EPA 1998; EPA 2001c.) For decades, the wastewater was either allowed to drain directly into the ground or, as in Sacramento County, pumped into abandoned gold-mining pits. (JAWA 1957.)

Following the fatal explosion in 1997 at American Pacific's Utah plant, a former employee at the old Nevada plant told the Las Vegas Review-Journal that in the early 1970s perchlorate tanks and pipelines, and spoiled batches of the chemical, were routinely dumped directly onto the desert floor. "Whenever they drained the 'chlorate plant, they just drained it on the blacktop and it drained into the desert. . . . They would [also] take 55-gallon containers of contaminated product and they would take them out and just leave them in the desert," said the former employee.

Only recently, and in only a few places, was technology installed to reclaim perchlorate from the wastewater stream. Even after reclamation, perchlorate concentrations in the wastewater stream remain hundreds of times higher than the EPA's most likely drinking water standards. No one can say for certain how many millions of pounds of perchlorate have been heedlessly flushed into the environment during the last half-century. Considering the amount of the chemical necessary to be detectable in Lake Mead and the Colorado River, the volume of perchlorate waste must be vast.

Perchlorate contamination of soil, as well as surface water and groundwater is being increasingly discovered at rocket testing and disposal sites (both military and private) where open detonation of rockets occurred. (EPA 2001a.) Surface water at one such site in Arkansas has been contaminated with levels of perchlorate reaching as high as 480,000 ppb.

The first public record of perchlorate contamination in California dates back more than forty years. In 1957, a report in the Journal of the American Water Works Association describes how "several California municipalities have experienced pollution of ground water supplies as a result of local underground disposal practices [of rocket fuel waste]." The article says potassium and ammonium perchlorate wastes flushed into groundwater in eastern Sacramento County had spread over several square miles, with perchlorate concentrations of 3.5 million to 5 million parts per billion in water. (JAWA 1957.) It wasn't until seven years later, though, that the first official state report of perchlorate contamination was released.

In 1964 the California Department of Water Resources tested groundwater in Folsom and eastern Sacramento. Among many other toxins detected, the state found perchlorate in 34 wells at levels of up to 18,000 ppb, with the highest concentrations on the property of the company then known as Aerojet General. (CADWR 1964.) However, the water department's report lists the results of perchlorate tests conducted as early as 1953, so the state has known of the potential for contamination at least that long. Despite these findings, the state declared the water supplies safe, and Aerojet's pollution continued.

In 1979, the California Central Valley Regional Water Quality Control Board found that perchlorate could be used to trace other chemical contaminants such as trichloroethylene (TCE) and perchloroethylene (PCE) as having come from Aerojet. At the time, tests of drinking water wells in Rancho Cordova found perchlorate at 220 to 300 ppb. Yet the cleanup of other toxins did not even attempt to address perchlorate. This was a major blunder, because the cleanup process for the other chemicals involved extracting toxic groundwater, partially cleaning it, and then reinjecting the water - still contaminated with perchlorate at concentrations of up to 8,000 ppb - into layers of groundwater which were previously clean.

Arden Cordova Water Service, the company that supplies drinking water to Rancho Cordova, later sued Aerojet for negligence and fraudulent concealment over this practice, alleging that the company "knowingly reinjected water still containing high levels of perchlorate," contaminating dozens of wells. The water supplier also sued the Central Valley water board and the state Department of Toxic Substances Control, saying both agencies had rejected its attempt to stop the reinjection. The suit alleges that the water board "willfully allowed" Aerojet to reinject the contaminated water into previously untainted wells, and that the Department of Toxic Substances Control kept from the public the extent of the contamination. Aerojet denies knowing that the reinjected water was contaminated.

In 1985 perchlorate reared its head again. This time the problem was at the Aerojet facility in Azusa, Los Angeles County, which had been designated a federal Superfund site in 1984 because of trichloroethylene (TCE), perchloroethylene (PCE) and carbon tetrachloride (CTC) contamination. Aware that Aerojet had handled millions of pounds of perchlorate on the site, the EPA's San Francisco regional office tested groundwater in the area and found perchlorate concentrations of up to 2,600 ppb in 14 wells. (EPA 1998.) According to sources in the Superfund program, EPA then notified local, state and federal authorities, but the U.S Agency for Toxic Substances and Disease Registry (ATSDR) felt that the method used to analyze perchlorate's toxicity was inadequate. EPA requested further guidance from ATSDR, but when none was forthcoming, turned its attention to other chemicals at the site.

By 1997, the California Department of Health Services (DHS) had developed a new analytical method that could reliably detect perchlorate concentrations in water as low as four parts per billion. New tests showed the extent and level of contamination at Rancho Cordova to be far greater than previously known - then found to be up to 6,000 ppb and later found to exceed 600,000 ppb. Other Superfund sites associated with Aerojet and Lockheed Martin were tested, and health officials realized they had a statewide problem. The other companies identified by the EPA as being responsible for perchlorate contamination of California's groundwater are Alpha Explosives, Bermite Powder Company, B.F. Goodrich, Grand Central Rocket, Boeing/Rocketdyne, United Technologies and Whittaker Corporation.

Perchlorate is not removed by conventional water treatment processes or those that remove chemicals such as TCE or PCE. The "cleanup" method most widely used today is to blend contaminated water supplies with uncontaminated sources - in other words, the discredited idea that "dilution is the solution to pollution." (EPA 2001c.) Nanofiltration and osmosis, which force water through membranes with extremely small pores, can remove perchlorate from water, but these technologies are extremely expensive, especially with such large areas of contamination involved. Biological treatment (using bacteria to convert perchlorate into a less toxic or innocuous compound) and ion exchange systems (replacing the perchlorate molecules with chloride) have been tested, but it is still somewhat unclear if they will work on a large scale. Ion exchange also produces its own toxic waste, a concentrated perchlorate brine which must be either disposed of or treated.

No matter which cleanup technology is used, the costs will be tremendous. Cleaning one part of the Aerojet Superfund site in Rancho Cordova, for example, is expected to cost $55 million. In documents filed with the federal Securities & Exchange Commission (SEC) in 2000, Lockheed Martin estimated the cost of complying with existing cleanup orders from the state and EPA as $140 million in San Bernardino County and $100 million in the San Gabriel Valley. (Lockheed 2000.) Even more startling is the estimated time frame for perchlorate cleanup. When the EPA recently unveiled its cleanup plans for the Rancho Cordova site, residents were stunned to learn that the agency's preferred plan would take an estimated 240 years. Aerojet, however, wants to take even longer - 348 years. SEC documents indicate that Aerojet doesn't expect to even determine the best cleanup technology for another 15 years. (SEC 2000.)

Over two to three centuries, and considering that an unknown amount of contamination has yet to be detected, costs for cleaning up perchlorate pollution in the nation's groundwater could easily reach billions of dollars. The final cost is unknown; what is known is that Aerojet, Lockheed Martin and other responsible parties won't be paying most of it.

When a company is liable for environmental contamination of a site where work was done under government contract, federal law allows the contractor to "recover" the cleanup costs from future government contracts - in other words, to pass a big chunk of the cost on to the American taxpayer. Aerojet's SEC filings report that of the company's current estimated liability of $353 million for cleanup of all contaminants at all U.S. sites, it will recover $213 million, or 60 percent, from the government and other parties who share responsibility. But its deal on perchlorate in California is even sweeter: In January 1999, Aerojet and the government reached a cleanup settlement covering both the Sacramento and Azusa sites whereby "the Government/Aerojet environmental cost sharing ratio was raised to 88 percent/12 percent from the previous 65 percent/35 percent." (SEC 2000.) Aerojet also stands to benefit even further from the government-financed cleanup: Once the water supplies meet health standards, the company plans to sell part of the site for commercial development. (Bowman 2001.)

Lockheed Martin, the contractor responsible in Southern California, has negotitated a similar deal, but the exact percentage of reimbursement is unknown. (EWG 2001.) According to the company's most recent SEC filings, "A portion of our business is classified by the government and cannot be specifically described," but:

Under an agreement with the U.S. Government in 1990, the ... groundwater treatment and soil remediation expenditures referenced above are being allocated to our operations as general and administrative costs and, under existing government regulations, these and other environmental expenditures related to U.S. Government business . . . are allowable in establishing the prices of our products and services. As a result, a substantial portion of the expenditures is being reflected in our sales and cost of sales pursuant to U.S. Government agreement or regulation. (SEC 2001.)

In addition, as a party that shares responsibility for Lockheed's San Gabriel Valley contamination, the Air Force has agreed to directly pay 50 percent of cleanup costs there. This is in addition to the costs that will be borne up front by Lockheed but later reimbursed by the government. Lockheed also states that as a contractor it runs a number of government-owned facilities where unspecified groundwater contamination has been discovered. Under the usual provisions of Superfund law, a company that shares responsibility at a site owned by another entity also shares the cost of cleanup, but Lockheed says: "At [government-owned] facilities, environmental compliance and remediation costs have historically been the [sole] responsibility of the government." (SEC 2000.)

The bottom line: At Lockheed-owned sites where the government shares responsibility, the public is paying half; at government-owned sites where Lockheed shares cleanup responsibility, the taxpayers are footing the entire bill; and of the cost that remains, the government will reimburse Lockheed for a large part.

Apparently those arrangements aren't good enough for the contractors. Both Lockheed Martin and Aerojet have been major contributors to a group of 188 political action committees which lobby Congress to repeal the federal Superfund tax on polluting industries. The tax generated $4 million a day to pay for toxic cleanups, including sites of perchlorate contamination, until 1995 when industry's influence won out and the tax was repealed. From 1991 to 1998, Lockheed contributed more than $1.5 million to get (and keep) the tax repealed; GenCorp, the parent company of Aerojet, also contributed more than $300,000 to these anti-Superfund PACs. (USPIRG 1998.)