The Environmental Protection Agency is reopening the debate on the health effects of polychlorinated biphenyls, or PCBs, as part of a court-mandated review of rules setting up a ban on the substance three years ago.

PCBs, used largely by utilities in transformers, capacitors and other electrical equipment, were the only chemicals singled out by Congress in the Toxic Substances Control Act of 1976 because they do not break down easily, can work their way into the food chain and have become widespread.

Because certain PCBs can produce liver tumors when fed to mice and some studies have indicated increased incidence of cancer among those exposed to PCBs, the compound generally has been thought to be a carcinogen. PCBs were recently listed in the Health and Human Services Department's second annual report on substances "either known or reasonably anticipated to be carcinogens."

But in a recent Federal Register notice, EPA said two new analyses--one by Drill, Friess, Hays, Loomis & Shaffer Inc., which was commissioned by the Edison Electric Institute, and another by Ecology and Environment Inc., commissioned by the Chemical Manufacturers Association--have revealed new evidence showing "that PCBs do not pose any serious risk of injury to human health."

Larry O'Neill, a spokesman for Monsanto, the only firm that was producing PCBs commercially when Congress passed the ban, said, "I think industry's general feeling is that since the Toxic Substances Control Act was passed, there have been additional health effect studies which have shown that PCBs, while not harmless, are not carcinogenic and do not have serious long-term health effects."

Edison Electric, which represents most major utilities, has asked EPA to allow companies to continue to use most equipment that contains PCBs. Under the Edison proposal, when that equipment wears out, some could be replaced with equipment that contains PCBs, but not all of it.

EPA, under court order to develop final standards by Aug. 19, is holding a symposium Wednesday and Thursday to examine the new health studies. The symposium has upset some environmentalists and government scientists because the panels include more industry representatives than independent scientists.

The critics argued that any symposium on PCBs should include more independent scientists involved in PCB research who could present papers on their work.

EPA spokesman Joe Handy said the purpose of the symposium is "to review, discuss and interpret PCB scientific data published since 1978 and preliminary data from ongoing studies. One reason the chemical companies are so involved is because they're doing most of the ongoing work."

Flares also went up recently when symposium participants were asked to sign a confidentiality agreement. The memo from Life Sciences Inc., which is under contract with EPA to organize the symposium, said, "Our customer has requested that you not disclose any information on this assignment until it is a matter of public record."

When a participant complained to Life Sciences, she was told there had been a misunderstanding.

"It's much ado about nothing," said Jack Glennon of Life Sciences. "She called us up and we acknowledged that it was our error." EPA declined comment.

The Natural Resources Defense Council has asked EPA administrator Anne M. Gorsuch to cancel the symposium, charging that EPA organized it hastily to develop a public record justifying a final rule suitable to industry.

Two years ago, EPA proposed a rule that exempted most existing equipment from the PCB ban, but the U.S. Court of Appeals for the District of Columbia sided with the Environmental Defense Fund, which had argued that the regulation was too weak and did not carry out the congressional mandate.

Congress had directed EPA to ban the manufacture and distribution of PCBs by 1979. It also prohibited the use of any PCBs in existing equipment, unless the PCBs were "totally enclosed."

In the original rule, EPA defined about 99 percent of the equipment containing PCBs as "totally enclosed," according to Jacqueline Warren, the EDF attorney who supervised the suit.

Although EDF won the case, it agreed with its opponents that an immediate ban would have severely disrupted electric service throughout the country. Instead a schedule for investigating and developing the stricter standards was set.

Since then, EPA has found that, based on industry data, large PCB capacitors do leak, to the tune of about 439,000 pounds of PCBs each year. In a proposed rule, EPA has recommended that the PCB capacitors be phased out over 10 years.

Warren, who is now with the NRDC, praised that recommendation but criticized the agency's failure to apply the same standard to transformers.

EPA estimated, based on industry data, that PCB transformers may leak 25,000 pounds of PCBs annually. It noted that because many transformers "are located in the nation's electrical distribution system, leaks from these transformers have the potential to expose humans and the environment to PCBs."

Nevertheless, EPA has proposed authorizing the use of PCB transformers "indefinitely," arguing that it would cost industry billions of dollars to replace the equipment.

EPA has proposed an inspection and maintenance program to reduce the risk of leaks or explosions, but Warren said accidents could still occur.