ASHINGTON, Jan. 3 — Three months before workers refueling an Ohio nuclear reactor discovered last year that its lid had rusted nearly all the way through, the staff of the Nuclear Regulatory Commission drafted an order to close it for inspection.

But the order was never issued, because the staff doubted its authority to close the plant, did not want to impose unnecessary costs on the owner and was reluctant to give the industry a black eye, according to an internal commission report released today.

The report, by the commission's inspector general, concluded that the staff had been too hesitant and that a policy adopted by the N.R.C. in the mid-1990's to take costs into account when setting regulatory requirements was in conflict with the commission's goal of maintaining reasonable assurance of public safety.

But the basic problem, the report said, was the staff's assumption about who had the burden of proof — the commission or a plant's operator — when safety was in question.

The commission "appears to have informally established an unreasonably high burden of requiring" of itself "absolute proof of a safety problem, versus lack of reasonable assurance of maintaining public health and safety," said the inspector general, Hubert T. Bell.

The report, dated Dec. 30, was issued today after an account about it appeared this morning in the Cleveland daily The Plain Dealer. Its sharp criticism of the commission's staff concerned the belated nature of the shutdown of the Davis-Besse reactor, near Toledo, last year.

Other reactors of the same design had been found to have cracks in parts attached to the lid, and the commission wanted all such plants inspected by Dec. 31, 2001. The operators of the Davis-Besse plant wanted to wait until March 2002, when the reactor was scheduled to be shut anyway for refueling.

When the plant finally closed, on a compromise date in February 2002, engineers and workers were shocked to find that cracks of the kind the commission staff had suspected there had let acidic water leak onto the head, where it had eaten away a 70-pound chunk of steel six inches thick.

Only a layer of stainless steel about a quarter-inch thick had prevented the cooling water from spewing out of the vessel head, in a leak that could have proved catastrophic. The corrosion was the most extensive ever found at an American nuclear plant.

Three months earlier, in November 2001, the commission's staff had drafted a shutdown order. But some staff members were not sure they had the authority to issue it, the inspector general's report found. Others thought that it might not be defensible in court, and that such an order would "destabilize confidence" in the industry.

William M. Beecher, director of the commission's public affairs office, said the N.R.C. received the report on Thursday and had not yet determined how it would respond. But, he said, "the N.R.C. has the unquestionable and unquestioned authority to shut down a plant if it concludes that public health and safety is potentially in jeopardy."

Such shutdown orders were common in the 1970's and 1980's, when reactors were newer and operating problems were first occurring. They are rarer now. In the mid-1990's, the commission adopted a policy called "risk-informed regulation," in which it pays more careful attention to the costs it imposes on plant operators, comparing those costs with the amount of risk reduction they provide.

But Mr. Beecher said that while the commission and its staff do take costs into account, "the primary and overarching requirement, concern, standard, for the N.R.C. is public health and safety."

"Anything else," he said, "is secondary or tertiary."

As for the concern about having to defend such an order in court, the new report determined that the fear of a lawsuit had been unfounded. The president of the FirstEnergy Nuclear Operating Company, which runs Davis-Besse, told the inspector general that no formal shutdown order would have been required; he would have closed the plant had the commission simply telephoned and asked him to do so, he said.

The inspector general undertook his investigation at the request of the Union of Concerned Scientists, a safety group that is generally highly critical of nuclear operations.

David Lochbaum, a nuclear engineer with the group, said in an interview that shutting down early for a special inspection would not have been an undue burden on Davis-Besse. Other reactors suspected of cracks in the lids did just that, Mr. Lochbaum said.