Early this year, the Energy Research and Development Administration assigned an official named Christian W. Knudsen to head a task force looking into supplies and production costs of natural gas into the 21st century.

Dr. Knudsen, a chemical engineer who came to the government from the Exxon Corp. three years ago, completed some preliminary calculations in early March.

Me made no big claims. Instead, he emphasized the the results rested, as they had to, on a series of assumptions, best-guesses and projections.

Still, those projections conveyed a message that jarred the conventional wisdom: the cost of producing 1,000 cubic feet of gas at wellheads in the continental United States - including all taxes and a return of 15 per cent - would average no more than $1 until the end of the 1990s.

In addition, his figures showed supplies up to the year 2000 to be far more abundant than the public has been led to believe.

More importantly, Kindsen's numbers ran smack in to the ominous supply/price picture painted by the oil and gas industy and by President Carter in his then-upcoming National Enery plan. Carter has proposed a price of $1.75 per 1,000 cubic feet for new gas.

In the ensuing behind-the-scense brouhaha, ERDA officials summarily threw out Knudsen's estimates.

And Knudsen was abruptly fired as chairman of the ad boc Intermediate Energy Working Supply Committee, the task force of about 30 persons he had been assigned to run in addition to his normal duties as assistant director for review and anlysis under fossil-fuels chief Philip C. White.

Although Knudsen declined to discuss the affair with a reporter, an account of it was pieced together in interviews with ERDA officials and other sources, including congressional investigators.

The central issue emerging from the interviews was not the eternally arguable validity of Knudsen's curves, but the existence of slippery slopes awaiting those in the energy bureaucracy who may somehow become tainted by heresy.

It began in January with a commandment from ERDA's acting administrator, Robert W. Fri, to Dr. White to do a study to help determine what the agency's research and development priorities ought to be, and do it fast - by April 7.

A critical component of the study had to be estimates of future natural gas supplies. If gas will be relatively abundant, the case for research and development on alternative energy sources obviously becomes less pressing than if gas will be in relatively short supply.

White, who went to ERDA in 1975 from the Amoco Oil Co., where he was vice president for research and development, initially picked I. Frederic Weinhold to head the task force, but replace him with Knudsen after Weinhold left to join the staff of White House energy chief James R. Schlesinger.

Even before Knudsen took over, the task force was getting indications that its result would be, in the word often heard in the interviews, "sensitive."

Without telling Knudsen, White quietly launched a parallel supply/price study by a second group.

Neither team had much of a problem charting the supply and production costs of proved reserves, or even reserves reasonably "inferred" to be available. Notably, they generally agreed that into the 1990s supplies will be relatively plentiful and production costs under $1.

But a major split did develop over the predicted availability and costs of producing so-called undiscovered reserves, but which, by definition, there is far more speculation than proved or inferred reserves.

Knudsen's group, relying entirely on public data from government agencies and industry sources such as the American Gas Association, ended up more optimistic than the parallel group, which relied on a model devised by the Standard Research Institute on the basis of some industry figures and premises classified as trade secrets.

In a phone interview, fossil-fuels chief White said the SRI premises were unavailable even to him. Knudsen was turned down when he asked Harry Johnson, director of the so called MOPPS (Market Oriented Program Planning Study) program of which the task force was a part, to let him see them.

On April 12 - Six days before President Carter would introduce a nationwide television audience to his NEP - White met with Hugh Guthrie, aide who had worked for Shell Oil for 33 years.

White, in the phone interview, sad Guthrie told him that the Knudsen curves would be "laughed out of court" by the industry and would jeopardize the credibility of the whole MOPPS effort in the eyes of John F. O'Leary, Schlesinger's top aide in the White House.

"I decided he was right," White said. Asked if he may have been influenced by political implications, he said, that none had "dawned on me at that moment."

Early in the interview, White scorned the "unreality" of the Knudsen curves, saying, for example, that Knudsen arbitrarily had figured the costs of producing undiscovered reserves at about $2.50 per 1,000 cubic feet, or approximately five times the 1975 regulated interstate price. Later, however, he said, "Hell, that's not an unreasonable guess."

He opted for the SRI curves as "more realistic," although he admittedly didn't know the premises from which they were derived. "It was a lousy choice I had to make" - under the April 7 deadline pressure - he said.

As a result of that choice, Johnson went to the White House to brief O'Leary on the SRI curves, which are more easily reconciled with the NEP. Yesterday, O'Leary told a reporter he's never heard of Knudsen, his curves or his methodology. He said "I'd like to be briefed" on them, but emphasized his belief that the facts, in any case, show that SRI curves to be "incorrect," that is unreflective of what he believes to be the reality: the nation is in an "attrition situation" insofar as the gas supply is concerned.

Last week, at a two-day meeting with industry experts in Reston, ERDA officials devised new curves that come closer to the dire estimates in the Carter plan. White said that all of the curves, including Knudsen's, will be published in a few days.

Before that, White had abruptly dismissed Knudsen, who had refused to compromise projections from public data to reconcile them with projections made from closed proprietary data. Neither man will say why. Some sources suggested that Knudsen was punished for heresy.

He is now back in his old job as an assistant to White.