Transcripts of the Nuclear Regulatory Commission's closed-door meetings during the Three Mile Island crisis portray an agency that, two days after the Pennsylvania nuclear accident, still lacked any clear idea of how to deal with the problem.

"We are operating almost completely in the blind," NRC Chairman Joseph M. Hendrie complained to his colleagues as he labored to inform Pensylvania's governor about the situation. "His information is ambiguous, mine is nonexistent. . . ."

Still, the transcripts show, the regulators, consulting continuously with White House public relations advisers Jody Powell and Gerald Rafshoon, worked hard to make sure that mainly "reassuring" information would reach the public.

"Which amendment is it that guarantees freedom of the press?" Hendrie asked his colleagues at one point." "Well, I am against it."

The transcripts also show that the commissioners had almost no confidence in the company they had licensed to operate the nuclear plant, that they had difficulty deciding whether or not to call for a mass evacuation from around the endangered site and that their own expert on evacuations had no knowledge of how people living near the plant might be moved to safer areas if necessary.

A review of the transcripts also makes it clear that the accident at Three Mile Island raised numerous questions that had not been contemplated when the reactor was licensed to operate.

"We saw failure modes," one technical adviser told the commissioners, "the likes of which have never been analyzed."

Concerned at one point about releases of radio-active gas from the plant, Pennsylvania Gov. Richard Thornburgh asked Hendrie, "Is there anyone in the country who has experience with the health consequences of such a release?"

Hendrie answered, "Ah, not in the sense that it's been studied and understood in any real way."

The NRC transcripts, the most striking behind-the-scenes portrait of government at work since the Watergate tapes, were made public yesterday by the House Interior and Insular Affairs Committee. The committee had forced the agency to release the transcripts by threatening a subpoena if they were not turned over.

The several hundred pages of transcript cover a fairly continuous commission meeting that began Friday morning, March 30, two days after the accident, and ran through Wednesday, April 4. Sessions Were Tapped

They were transcribed from tapes made whenever at least three of the five commissioners were present. The commission members seem to have known their sessions were taped, but they apparently did not know the transcripts would become public so soon.

The commission, which said the transcripts were in rough draft form, added a note indicating that commission meetings are normally recorded, apparently for transcript purposes. The recorders picked up most of the talk around the table and telephone conversations that were amplified over a speaker telephone.

The documents reflect an informal, confused series of gatherings, interrupted almost constantly by telephone calls and reports from commission staff aides.

The only time in the transcripts when the commissioners devote sustained attention to one subject is when they debate, sometimes for hours, the wording of news releases on the situation.

When the commissioners first came together Friday morning, they were faced with a growing hydrogen bubble in the Pennsylvania nuclear plant that threatened to drive out cooling water, bringing on the risk of a meltdown.

It is clear from the discussion that neither the commissioners nor their expects on the scene had any clear idea of how to eliminate the bubble. No Solution Seen

"We don't have a solution," Roger Mattson, the agency's director of safety systems, said over the long-distance telephone line from the site. "We've got the best you got, Joe, and they are not coming up with answers."

Late Friday evening, after the commissioners approved a news release suggesting that the problem was under control, another technical aide called to say that "the flash of inspiration that they had a couple of hours ago doesn't seem to be jellying, and they're still working, but we don't have anything further on it."

One difficulty was that the NRC officials did not trust the technical staff of Metropolitan Edison Co., the utility licensed to operate the plant, to handle the situation.

"Met Ed is not all that strong technically," Hendrie said at one point. He noted that the firm had support from engineers employed by its parent corporation, but later said that "the whole shop there is not that strong, and the operating staff clearly is not."

Harold Denton, the chief NRC official on the scene, shared that view, complaining that the utility had been far too slow to respond to the crisis.

"We kind of had the feeling this morning," Denton told the commisioners by telephone that Friday, "the licensee doesn't even recognize the problems that we're facing with regard to the bubble and damage and what might happen if we were to lose vacuum and so forth." Evacuation Argument

Mattson, the safety director, was so concerned about the risk of a melt-down that morning that he argued forcefully for a mass evacuation.

"We have got an accident that we have not been designed to accommodate," he said, "and I don't have a reason for not moving people. I don't know what you are protecting by not moving people."

Mattson's repeated warnings started the commissioners talking about how an evacuation might work. After a long-search-during which various staff members said, in effect, "it's not my job"-they finally found the official responsible for evacuation plans."

"Do we have some idea of precisely what would happen today?" asked Commissioner Victor Gilinksy. "Do they have places to tell people to go?"

The evacuation expert, Donn Collings, had no answer. "I imagine that they do," he said, "but you know, those are things that people decide when they do it."

The commissioners then asked Collings to "interface with whoever your counterpart person is" in Pennsylvania, and ascertain more definite information. Collings returned a half-hour later, saying a total evacuation could be carried out in about one hour.

"Let me ask you," Gilinsky said, "are you talking about Harrisburg, too?" The transcript then reads:

"Mr. Collings: 'Let's see, Harrisburg is in which country?' (Mumbling to himself, obviously looking at a map." Casual Information

The documents show that some information about the accident reached the NRC in casual ways.

"A second thing happened this morning," Mattson reported Friday. "Just in the midst of taking some data from the [utility] guy, he says, 'You know, I have just heard in the control room that a guy was reading the script chart recorder and says he's got a funny blip in the containment that nobody saw before.'

"That's possibly a hydrogen explosion . . . and that would be the source of this bubble," Matson added.

While they were still trying to determine the status of the reactor, the commission turned to the problem of how to deal with the media.

Hendrie had a series of telephone calls with Powell on this matter, and Hendrie indicated repeatedly that it was important to come up with a statement that would satisfy the White House press secretary.

Hendrie also passed on to his colleagues some advice Powell gave him about dealing with reporters. "He says watch out. . . . There will be a tape from [Denton] . . . at Three Mile Island and a tape of the dumb chairman answering the same question . . . and they will pick out the divergencies even though they may not be matters of substance."

Late Friday the NRC members began arguing about the text of a news release covering the day's events. Commissioner Richard Kennedy, who worked hard to tone down the release, said, "The focus, I think, has to be reassuring . . . reassure people that at least we're working on it."

While Kennedy wanted to eliminate most references to danger from the plant, Commissioner John F. Ahearne argued that the release might be misleading because "the whole flavor is very optimistic."

Ahearne argued that "it would be technically better if you said something about there's a possibility-it's small, but it could lead to serious problems."

Well, you've got to remember," Kennedy replied, "that's the only thing out of the press release they will then take out." 'Imminent Danger'

Eventually the commissioners settled on wording that there was no "imminent danger" at the plant, and they moved the information about the press release of radioactive gases from the first paragraph in a draft version to the last paragraph of the final release.

Two days later, another commissioner had second thoughts about the tone of the agency's public statements. A staff technician was explaining that it was still unclear how much pressure had built up in the containment vessel. "It sounds like the explosion is going to be a lot worse than we've let on," said Commissioner Peter A. Bradford.

The possibilities of a rupture of the containment - vessel prompted considerable discussion. Kennedy, for example said a "rough analysis" indicated there would be no rupture.

But Bradford said, "Joe [Hendrie] thinks it would. One rough analysis against another or at least Joe thought it might." Concern About Criticism

The members also expressed concern about outside criticism of the of the situation at Three Mile Island, principally remarks by Dr. Ernest Sternglass, a University of Pittsburgh physicist, who had stressed that nearby residents should have been evacuated.

NRC staff member Stephen Hanauer said, "Let's remember that most of the data Sternglass quotes is for chronic exposure, whereas this is one time and so this is less of an insult. On the other hand, if you get into a p--- contest with Sternglass, he's going to do it all over you because he's not restrained by [inaudible] anything that's going on."

Those involved in the discussions reflected several times on the possibility of the tapes of their meetings being made public. At one point, Kennedy urged that a conversation with Denton not to be released because Denton had criticized Met Ed's technical ability.

Hendrie said it would be "awkward" if the tapes were made public. but later observed that, "I suppose somebody will want to know, did we behave ourselves in this particular incident."