The Nuclear Regulatory Commission now believes that the uranium fuel in the Three Mile Island nuclear plant was completely without cooling water for as long as 50 minutes on the day of the accident, and that it was during those minutes that the fuel suffered most of its damage.

"The only thing cooling the fuel rods during that time was a very low flow rate of steam," Harold R. Denton, director of reactor regulation at the NRC, told the Senate subcommittee on nuclear regulation yesterday. "It's still too early to say how close we were to a meltdown but the core clearly reached very high temperatures, having come uncovered several times."

Denton said the early warnings of extensive fuel core damage were all there on the day of the accident (Wednesday, March 28) but that a combination of ignorance and instrument misreadings led the NRC to midjudge the extent of the damage.

A small explosion of hydrogen gas inside the concrete containment surrounding the damaged reactor took place at 1:50 p.m. Wednesday but the NRC wasn't told of it until 10 a.m. Friday. Had the NRC been informed of the explosion when it happened, Denton said, it would have known that large amounts of hydrgoen ahd escaped from the reactor to the containment-hydrogen that could come only from extensive damage to the core.

"Why would an operator withhold information like that for 36 hours?" Sen. Gary Hart (D-Colo), chairman of the subcommittee, asked Denton.

Replied Denton: "I don't think we know the answer to that yet." And NRC Chairman Joseph M. Hendrie said: "It was one of the most disturbing elements of what went on. It would have been enormously helpful to know as soon as that happened."

Another sign of heavy core damage was the level of radiation in the containment, but the NRC conceded yesterday that it chose to believe one set of instruments reading low radiation levels on the floor instead of another set reading high levels on the ceiling.

The floor readings were in the tens of rems while those in the dome were 800 rems. But the instruments in the dome are shielded from radiation so that technicians had to extrapolate to get the real radiation level, said the NRChs Roger W. Mattson. Extrapolation from 800 gave technicians a figure so high they didn't trust it, he said.

"That reading would have been thousands or tens of thousands of rems." Mattson said. "It was only later on that we calculated the dose inside at thousands of rems, all the way through the containment, not just the ceiling."

Meanwhile, at a separate hearing, NRC staffers told the Advisory Committee on Reactor Safeguards that the Three Mile Island core is so badly damaged that it looks like an inverted pyramid of debris and fragmented fuel that reaches down five feet along the 12-foot-long fuel rods.

Water flow throught he core is 93 percent blocked by the debris, and resistance to the circulating water is 200 times what it normally is. Water is flowing at "fractions of a foot per second," according to NRC Staffer Gary Holahan. That was still enough yesterday to have cooled the core down to the average temperature of 179 degrees Fahrenheit, the lowest it's been since the accident.

The NRC's Carl Berlinger said there are now 425,000 gallons of waste water contaminated with radioactivity on the floor of the containment basement. The water is five feet deep, almost twice the depth of two weeks ago. Berlinger blamed continuing leaks for the rising water levels.

At the Senate hearing, NRC people were repeatedly asked about the competence of the utility, the Metropolitan Edison Co., in dealing with the accident. In its replies, the NRC questioned Met Ed's technical competence no fewer than three times.

"We did not have from the very first day," Denton said, "the kind of things we needed to get from the licensee to judge what had occurred and what to do about it. This utility was not prepared to cope with this kind of accident."

Denton was asked if Met Ed's lack of competence implied a "systemic" problem in the electric power industry and he answered that many utilities had top engineering staff sbut that others did not. Those "at the lower end contract their engineering competence outside. That's difficult to deal with when you have an emergency."

At that, Sen. Hart responded: "At its worst, we have a 19th century industry running a 21st century technology."

Both hearings brought out the fact that an @NRC inspector had visited Three Mile Island twice in March, once for four days (March 19 to 23) and the second time (March 26) for one day.

The auxiliary pumps that were "blocked out" the day of the accident in violation of NRC rules were not closed off on March 22 but were apparently shut on the 26th during a test and were certainly shut on the 28th, the day of the accident.

"If in fact they were shut for two days," the NRC's Bruce Wilson told the Advisory Committee, "that's five shifts [of operators] and it's one of the largest mistakes." The NRC hinted the Senate hearing that it may deserve some share of the blame, since an inspector was in the plant on March 26.

"On a planned inspection like this, we allocate 20 percent of our time to look for things that are wrong," said the NRC's John David. "Unfortunately, in this case, nothing was seen." CAPTION: Picture 1, HAROLD R. DENTON . . . core reached " high temperatures"; Picture 2, SEN. GARY HART . . . questions operator's response