Bill Wade, Post-Gazette
	CMU graduate student Christine Wang examines a sediment sample. Wang and biological sciences professor William Brown, right, are using DNA fingerprinting techniques to identify which of the millions of bacteria in sediments are involved in breaking down PCBs. At center is Edwin Minkley the director of the Center for Biotechnology and Environmental Processes at CMU.

More than a quarter of a century after Congress banned the manufacture and sale of PCBs, people are still trying to find a way to make these pernicious chemicals disappear.

"We haven't found the magic pixie dust," said Edwin Minkley, a Carnegie Mellon University microbiologist. But a multidisciplinary team of CMU scientists and engineers is gathering clues that might someday provide new options for reducing the health risks associated with these manmade chemicals.

The researchers are presenting their latest findings in a series of talks that began yesterday and continue today at the American Chemical Society's annual meeting in Philadelphia.

For now, dredging PCB-contaminated sediments or burying them under layers of sand are the primary forms of remediation.

But by studying contaminated sediments from New York's Hudson and Grasse rivers, the CMU researchers have begun to identify families of bacteria that can detoxify and destroy PCBs. It might be possible to accelerate this bacterial action if they can determine the right mix of nutrients and sediment conditions.

And, in a demonstration this spring on the Anacostia River in Washington, D.C., environmental engineers laid a coke-filled carpet on the river bottom to bind up PCBs that otherwise might escape from the sediments into the water. Future versions of this "reactive core mat" might be designed not just to grab hold of PCBs, but to rip them apart.

"Even if it takes [the PCBs] a hundred years or 200 years to degrade, the cap above them gives you at least a hundred years" of protection, said Greg Lowry, an environmental engineer who is developing the mat technology.

For now, however, it's easier to say what conditions are not conducive to detoxifying PCBs than to say how that activity can be enhanced, said environmental engineer Jeanne VanBriesen.

"The complexity of the system has really surprised us," she said.

From benefit to liability
PCBs, or polychlorinated biphenyls, were once popular in a number of industries because of their seeming indestructibility. They are chemically stable, won't burn, don't conduct electricity and boil only at extreme temperatures. They were used as insulating fluids in electrical transformers, as plasticizers in paints and plastics and in pigments, dyes and carbonless copy paper.

Chemically speaking, the molecules consist of two benzene rings ---- each containing six carbon atoms ---- to which are attached up to 10 chlorine atoms. By varying the number and placement of chlorine atoms, chemists could make 209 different varieties of PCBs and each application typically included a mix of several different PCBs.

But when people realized that PCBs caused cancer in animals, and probably caused human cancers as well, that indestructibility became a liability. PCBs also can harm the immune, reproductive, endocrine and nervous systems.

Significant amounts of the 1.5 billion pounds of PCBs that were produced for five decades ended up in sewers or otherwise found their way into lakes and rivers. PCBs aren't soluble in water, so they tend to stick to whatever particle they find. "Whatever happens to the particles, that's the fate of the PCBs," said David Dzombak, an environmental engineer who heads the CMU effort. Many end up mired in sediments.

If that's where they stayed, people wouldn't be in much danger. But due to processes still not fully understood, some PCBs are released from the sediment and find their way into the food chain. Though insoluble in water, PCBs are soluble in fat and accumulate over time in fish, where they can reach dangerous concentrations.

Consequently, health officials recommend either limiting or completely avoiding consumption of fish caught in many urban rivers, including Pittsburgh's Three Rivers.

Because PCBs are so chemically stable, no one can be sure how long they might persist in the environment. Dredging of contaminated sediments has thus been considered the only sure way to reduce, if not eliminate, PCB levels.

That's an expensive proposition. General Electric, a major PCB user, is facing costs estimated at $500 million or more to dredge and backfill 40 miles of the Hudson. U.S. Steel in 1998 agreed to spend $30 million to remove 700,000 cubic yards of sediments containing PCBs and other hazardous wastes from five miles of the Grand Calumet River near its Gary, Ind., Works. On Friday, the company committed another $21.5 million to help cleanup several more miles of the river

Betsy Mallison, a spokeswoman for the state Department of Environmental Protection, said the department monitors Pittsburgh's rivers for PCBs, but has no current plans for remediation.

Figuring out deterioration
Though PCBs are long-lived, GE researchers uncovered evidence more than 20 years ago that they do deteriorate in sediments, said Minkley, director of the Center for Biotechnology and Environmental Processes in the CMU biological sciences department. And, in 1982, microbiologist James Tiedje of Michigan State University showed that bacteria are responsible for this degradation.

This raised the possibility that PCBs might be treated without dredging.

"We don't think dredging is necessarily the right way to go," said William Brown, professor of biological sciences, noting dredging releases some PCBs into the water.

He and graduate student Christine Wang have been using DNA fingerprinting techniques to identify which of the millions of bacteria in sediments are involved in breaking down PCBs. The hope is that learning the care and feeding of these bugs might yield new ways of treating PCBs without dredging.

But the information might be useful even if dredging is necessary, he added. "If we're going to take [sediment] out of the river, let's do something to get rid of the PCBs and make that soil available for use again," he explained.

The researchers also are trying to understand why PCB degradation occurs faster in some rivers than in others and why some areas show little, if any, evidence of degradation.

It's still impossible to say how long it takes to degrade PCBs, Minkley said. In lab experiments using sediment samples from the Grasse and Hudson rivers, fresh PCBs added to the Hudson sediments are dechlorinated in about seven weeks, compared to several months in the Grasse sediments.

Grasse sediments are more sandy than the clay-rich Hudson sediments, Brown noted. But there likely are also differences in the mix of nutrients in each river, and in the so-called "bioavailability" of those nutrients. The only thing that seems clear is that the difference in degradation rates is not related to differences in bacteria, he noted.

The dechlorination rates in the lab experiments can be misleading. The experiments involve "fresh" PCBs, not the aged PCBs that have been in the river for years and have attached to various particles, thus becoming less bioavailable, Brown said.

Minkley said it probably takes decades, though not centuries, for bacteria to detoxify PCBs in sediments.

VanBriesen said the CMU work has shown that the conditions necessary to totally degrade PCBs are unlikely to exist in any one place. The removal of chlorine atoms -- which drastically reduces the toxicity of PCBs -- is done by bacteria that work in the absence of oxygen, deep within sediments. Destruction of the remaining benzene rings, however, requires bacteria that use oxygen -- lots of it -- and are found in the upper reaches of the sediment and in the water.

That will make remediation without dredging difficult to engineer, she added.

Putting a cap on it
One technology that might at least buy some time for natural processes is called "capping." This has usually involved adding a layer of sand over the sediment, increasing the distance that PCBs and other hazardous substance must travel to escape into the water, said Danny Reible of the University of Texas, co-director of the Hazardous Substance Research Center/South & Southwest.

	Online Chart: Factors that release PCBs fromsediment

That works well in some situations, but not all. So engineers have designed a number of "active caps" which either do more to separate water from sediment, or include materials that can immobilize or destroy pollutants. Reible organized a project to demonstrate several of these caps in a stretch of the Anacostia River near the Washington Navy Yard this spring.

One such cap, developed by Lowry, is a half-inch-thick layer of coke breeze -- a byproduct of the coke used for steelmaking -- sandwiched between layers of geotextiles. This blanket is produced in sections that are 10 feet wide and 100 feet long; each blanket is unrolled like a carpet on the river bottom and then covered with six inches of sand.

In the Anacostia demonstration, 10,000 square feet of the river bottom was capped, using $950 worth of coke breeze from the Clairton Coke Works. PCBs bind about 100,000 times more tightly to the coke than they do to sediments, Lowry said.

The mat also makes it possible to distribute more expensive materials over a large area, he said. So it may be possible to also sandwich substances such as reactive iron, which can help degrade PCB or other hazardous wastes.

Capping and remediating PCBs where they rest isn't particularly popular with river stakeholders. "They just want it removed," Lowry said. But even when dredging occurs, some residual PCBs remain. The Hudson dredging project, for instance, includes plans for backfilling, which is another word for capping, he added.

Just three years into the multidisciplinary PCB project, which is funded by the David and Lucile Packard Foundation, the CMU researchers haven't come up with any solutions yet, Dzombak acknowledged.

"It's complicated," he said, "but we think we've made some progress."