The U.S. Atomic Energy Commission's study on reactor safety seriously underestimates the hazards of nuclear reactor accidents, the Cambridge-based Union of Concerned Scientists and the Sierra Club said in a joint report presented to the AEC on Friday.

The report, prepared by Henry W. Kendall, professor of Physics at MIT and Sidney Moglewer, an operations research analyst with the McDonnell Douglas Corporation, charges that the risk of a reactor accident in the early 1980s is over 400 times greater than that predicted by the reactor safety study, which was released in August.

The safety study, financed by the AEC and conducted by a group of independent researchers led by MIT professor of nuclear engineering Norman C. Rasmussen, concluded that nuclear power plant accidents present a negligible risk to public health and safety.

James J. MacKenzie, chairman of the Union of Concerned Scientists, yesterday attributed the two reports' different conclusions to the methods of safety analysis used in the AEC study.

MacKenzie claimed the AEC report's techniques have been proven unreliable in their failure to estimate accident probability in the aerospace industry.

In-Home Study

David R. Inglis, a University of Massachusetts-Amherst professor of physics who worked on the Union-Sierra Club report, said yesterday the reports' conclusions differ because the AEC report was written by people "whose bread and butter comes from the nuclear enterprise."

"We've gotten closer to the truth by not being deluded by the mass psychology of the nuclear power enthusiasts," Inglis said.

In their report, the Union Scientists and the Sierra Club also charged that the AEC is "misusing" the results of the Rasmussen report.

Whenever the AEC study has been summarized for public dissemination, the scientists said, "restrictions, limitations and uncertainties have all dwindled and vanished."

Congress will consider the AEC study when it debates the renewal of the Price-Anderson Act. The Act, which expires in 1977, provides for the Federal government to assume limited liability for reactor accidents.

Donald F. Ford '70, executive secretary of the concerned scientists' union, said yesterday the postponement in considering the Price-Anderson Act until the next session of Congress indicates recent criticism of the AEC study has eroded support for the Act's renewal.