A Nuclear Regulatory Commission memo warned in January last year of what became a major cause of the Three Mile Island accident, but the memo never made it out of the commission to the nation's operating utilities, a House subcommittee was told yesterday.

Instead, according to NRC officials later, "the problem was reviewed within the agency and dismissed."

The memo, signed by Thomas M. Novak, chief of the NRC's reactor systems branch, warned that in some cases the level of the cooling water in the radioactive core of a reactor might seem higher on some instruments than it actually was. That was precisely what happened March 28 at Three Mile Island, where operators relying on a particular pressurizer gauge allowed the core to become uncovered.

The memo was directed to about 15 other members of Novak's unit for their guidance in reviewing applications for construction permits and new reactor licenses. It advised them to check applications for ways around the problem.

"This is the first I've seen of this," said John G. Herbein, vice president of Metropolitan Edison Co., which owns Three Mile Island, when presented the memo at yesterday's hearing of the House energy and environment subcommittee. "To my knowledge we were not apprised of this memo."

Rep. James Weaver (D-Ore.), who chairs the subcommittee's task force investigating Three Mile Island, said he was "very disturbed . . . shocked" that the memo had been written 30 days before Three Mile Island II received its operating licenses and yet had not gone to the utility.

Roger Mattson, director of the NRC's system saftey division, said in an interview: "In those days, the information seemed more theoretical than real. It was a far-out possibility in the minds of men that looked at it . . . the problem was reviewed within the agency and dismissed."

In the wake of Three Mile Island, he added, "the memo was important, but hindsight is always clearer than foresight."

Although Novak signed the memo, it was written by a staff member, Sanford Israel, in the aftermath of a September 1977 accident in which the Davis Besse plant near Toledo, Ohio, suffered a loss of cooling water, Mattson said. About the same time, Israel received a handwritten draft copy of another report on the same subject, this one by an NRC consultant named Carl Michelson.

The Michelson draft, however, also discussed other problems and, according to Mattson, it was not a factor in Israel's final memo. Michelson's draft also circulated among members of the NRC's technical advisory board and the problem was mentioned in construction permit talks with Portland, Ore., General Electric Co. but was not given any particular urgency, Mattson noted.

The pressurizer gauge difficulty eventually became the subject of an exchange of letters between Michelson and the Babcock & Wilcox Co. that built Three Mile Island, but nothing had been resolved when the accident occurred March 28.

Rep. Thomas J. Huckaby (D-La.) told industry witnesses at the hearing that "it just blows my mind" that no direct, reliable measurement exists of water level in the reactor core. Installing such instruments on all existing reactors would cost" a measly $10 million," he said.

Babcock & Wilcox vice president John H. MacMillan responed that in the company's view operators have sufficient information to know the condition of the core water and to recognize problems with the pressurizer if they will only pay attention to it.He noted that the only mechanical failure at Three Mile Island involved a stuck-open pressure valve that allowed water to flow unimpeded out of the hot reactor core.

Three Mile Island station manager Gary P. Miller, who was in charge of efforts to control the situation March 28, told the hearing he had a great deal to attend that day. "The phone, the pressure, the fact that the plant was in a state that I had never been schooled in, combined to make conditions almost intolerable," he said in his written testimony. "However, the control room remained calm," he added.