The Nuclear Regulatory Commission is suffering from an identity crisis.

In the wake of Three Mile Island, the nation's closest brush with nuclear disaster, self-criticism has reached the proportions of cultrual revolution within the agency responsible for keeping watch over nuclear power.

"the mythology of nuclear power is crumbling the Three Mile Island was part of the pattern," said Commissioner Peter A. Bradford.

The five commissioners agree on very little, but they were unanimous in separate interviews on one point: their debating-society style, however, they have yet to begin discussing what to do about it.

At very level of the four-year-old commission, the traditional trust in nuclear power and in the nuclear industry's electric utilities is being reassessed.

"up to Three Mile Island, the perception was that we were dealing with basically a responsible industry that shared our concern about these dangerous machines and [that] could be relield on to do what was needed," said Robert Minogue, chief of the NRC's office of standards. "now we're questioning that."

The NRC also is questioning itself.

"the spectrum used to range from sanguine to alarmed" about the safety of nuclear power and the ability of the agency to protect public health, said a top-level staff member. "Now the whole spectrum has moved over a notch. The sanguine are now worried and the alarmed are scared to death."

The 2,070 scientists, engineers, administrators and technicians that work at the NRC had a curiously mixed reaction to Three Mile Island. The highest paid collection of bureaucrats in the government (average salary: $34,686), boasting 251 doctorates, they had become, as they now see it, complacent.

"we had never even heard the words 'crisis management,'" said one division manager, "because nobody ever really thought there would be a crisis."

When a series of mistakes equipment failures and quirks of fate uncovered the Three Mile Island reactor core March 28 and sent raioactivity billowing out over Pennsylavania, the NRC system response was "miserable, utter chaos," the manager said. Other used words like "panic," "total disbelief" and "pandemonium."

But, they insists, individuals overcame their system and the incident was manhandled into submission. "It made everybody proud, but it scared us all at the same time," said Roger Mattson, director of systems safety, who now heads an NRC "lessons learned" task force of 100 persons working full time on the accident. "a lot of people are still losing sleep over it".

The change in NRC attitudes is a fundamental one. It has been under way since the agency was created out of the old Atomic Energy Commission in 1974, but Three Mile Island accelerated the process.

Alone among the activities that the federal government regulates, nuclear power was created by the federal government. The AEC was openly out to promote nuclear power, to tame the atom for peacful use and to make electricity too cheap to meter. Atomic power was a new, cheap way to boil water.

"there has always been this feeling that nuclear power is fragile, that it is so complex, so beautiful, so delicate that it has to be protected from people who don't understand it," said Harold Lewis, a physicist to the University of California in Santa Barbara.

"everywhere there was this hyperbole," said Commissioner Bradford. Because of a fundamental belief in the basic concept, the scientific community assumed without much probing that a little radioactivity never hurt anybody, that the waste problem would be solved easily, that plutonium by-products were too hot to handle for terrorists or underdeveloped countries looking to make bombs and that a major nuclear power plant accident was wildly improbable.

"we felt a sense of mission, that we're in the royal service . . . guarding national security and the public trust," said another staff chief. With the AEC more interested in weapons than power plants staff decisions routinely were rubber-stamped.

One by one the assumptions collapsed.

Prodded by growing criticism of nuclear power, Congress abolished the AEC in 1974. Its regulatory functions and most of the minuscule regulatory staff went over to the new NRC; the public relations effort went to what is now the Department of Energy. Congress also abolished the Joint Congressional Committee on Atomic Energy, which had enthusiastically pumped out the money for AEC projects and defended nuclear power from all criticsm. Those days were over.

"the demise of the Joint Committee was probably the most significant part of this shift," said Commissioner Victor Gilinsky, former head of the Rand Corp.'s physical sciences department. Jurisdiction over nuclear matters was split among several subcommittees on both sides of Congress, many of them peopled by nuclear critics.

Gilinsky, the only member of the first NRC who is still on the panel, was an early skeptic on many nuclear issues. Routinely outvoted 4-1 in the early days, he has gradually seen the tallies shift to the point where he is occasionally in the majority.

But with all the old assumptions under fire, the NRC is now essentially leaderless.

There is deep division in fundamental outlook among the five commissioners. Gilinksky and Bradford, a former Maine public utility commissioner known for his consumer advocacy, tend to adopt a more critical, less trusting approach to the electric utility industry than do Commissioner Richard T. Kennnedy and Chairman Joseph Hendrie.

Hendrie, for example, the only commissioner to come up through the licensing ranks of the NRC and the only one with deep knowledge of nuclear reactors, voted with Kennedy against issuing formal orders to require utilities to make certain changes in the wake of Three Mile Island.

They argued that staff directives were enough of a prod of the industry. Kennedy, a retired Army colonel and former director of the National Security Council planning staff, said in an interview that "business as usual" should continue at the NRC, because the commission has always paid close attention to safety matters.

The swing vote is the newest commissioner, John F. Ahearne. He agreed with Gilinsky and Bradford that formal orders were needed to establish commission firmness, but earlier voted with Kennedy and Hendrie to delay taking action to shut plants similar to Three Mile Island. His background includes program analysis work in the Department of Energy posts and a stint with the White House Energy Policy and Planning Ofice.

"yes, things move slowly . . . it's part of the basic 'deliberative body' concept of regulatory commissions and partly the fact that the relationship between the commissioners and the (staff) directors has never been spelled out," he said. "it makes it hard to manage a crisis."

The staff directors are in charge of groups organized to compete with one another. Few below that level know what the other branches are doing, which results in duplication of effort and some backbiting. In the absence of any clear direction from above, technicians tend to delay making recommendations to achieve a united front against the clashing commissioners.

When Hendrie speculated during Three Mile Island that oxygen and hydrogen might combine to form an explosive bubble inside the reactor core, the staff was asked to figure out how much oxygen would make it a liable to detonate.

"nobody asked the basic question, which was not much oxygen was needed but rather whether any oxygen could escape at all," said a staff member involved in the debate. "It couldn't." The explosion could never have occurred.

The NRC's public worry about the chances for a blast, however, made the Three Mile Island incident an international crisis. This error is seen now as the worst of the episode, and engineers all over the NRC are kicking themselves that they didn't catch it sooner.

The NRC has not been allowed to carry on its self-examination in seclusion. "We're being investigated and questioned by representatives of every grouping of human beings in the western hemisphere . . . and it just cuts bloody hell out of our ability to get any work done," Hendrie complained. One industry executive counted 20 separate investigations into the events surrounding Three Mile Island.

Between appearances before investigators, commission members try to keep up with daily demands and worry a little what nothing much has changes in the way they argue for hours before reaching any decision.

Chairman Hendrie agreed that the structure was troublesome but said it was "premature" to propose a cure.

"if (a nuclear accident) happens tomorrow, I think I'll have better idea what to do with it, but I can't be more specific than that," he said. CAPTION: Picture, Copter sampling radiation over Three Mile Island after accident that has generated intense self-examination at regulatory agency. AP