The Department of Energy has greatly downgraded its controversial estimates of how much the federal government has subsidized nuclear energy over the years and in effect is arguing that the price of nuclear power is competitive with coal- and oil-fired electricity. This is a reversal of the department's previous position.

The 30-year total is not $40 billion as estimated last year but only $12.8 billion, a 68 percent cut, according to the department's revised report. The reduction, which the author of the original report charges was forced by the nuclear industry and pro-nuclear advisers to Energy Secretary James Edwards, was achieved primarily by not counting subsidies for military and future technology research.

The original report, compiled last year by economist Joseph Bowring before he left DOE's Energy Information Administration, said that federal aid over the period had kept the cost of nuclear power at about half what it would otherwise have been.

The rewritten study disagrees, saying subsidies have saved the nuclear industry only 0.47 cents over the current 2.25 cents per kilowatt hour that atomic power costs on leaving the plant gate. Bowring calculated the subsidy four to six times higher, at 1.6 to 2.5 cents. Adding only 0.47 cents would not hurt nuclear power's competitive position, but the higher figures would.

Bowring left DOE last October, before the study was formally complete. "Substantial revisions in the draft were made in response to reviewers' comments," the final report said.

The major change involved research and development subsidies, which Bowring listed as $23.6 billion and the final report tallied at $7.3 billion. Nuclear industry spokesmen had argued when Bowring's draft was made public that this was its major problem.

David L. McNicol, Bowring's former boss and now deputy assistant administrator for applied analysis, said the final total did not include $8 billion in research and development spending for hydrogen fusion reactors or the breeder reactor, which produces more fuel than it burns. "These are technologies of the future," and have not contributed to the existing domestic nuclear power industry, he said.

Another $1.9 billion in military research and development aid was excluded, along with $2.5 billion in pre-1966 biological and medical research on grounds it was primarily military. Safety research totaling $1.3 billion by the Nuclear Regulatory Commission, which Bowring included on grounds that safety rules would have had to be drawn up by the industry if the government had not done it, was also cut out.

The remaining research and development spending from the last 30 years was projected over the next 20 years rather than being counted as money already spent. This, McNicol said, was because capital expenditures are normally spread over time.

In addition, the final report excluded $2.5 billion in uranium market promotion, $2.8 billion in aid to fuel enrichment for foreign plants and $200 million in international programs, on grounds that these were only indirect subsidies to the domestic industry. Costs of the future disposal of spent fuel rods now being stored at reactors nationwide, the cleanup of abandoned uranium mill wastes and future reactor decommissioning were also eliminated, a total of between $3.3 billion and $4.3 billion, on grounds that industry will eventually shoulder the costs of research already done in these areas.

Bowring described the result as "garbage."