When he retired after 26 years as an investigator with the Nuclear Regulatory Commission's Office of the Inspector General, George Mulley thought his final report was one of his best.

Mulley had spent months looking into why a pipe carrying cooling water at the Byron nuclear plant in Illinois had rusted so badly that it burst. His report cited lapses by a parade of NRC inspectors over six years and systemic weaknesses in the way the NRC monitors corrosion.

But rather than accept Mulley's findings, the inspector general's office rewrote them. The revised report shifted much of the blame to the plant's owner, Exelon, instead of NRC procedures. And instead of designating it a public report and delivering it to Congress, as is the norm, the office put it off-limits. A reporter obtained it only after filing a Freedom of Information Act request.

The Fukushima nuclear disaster in Japan has thrust the NRC's role as industry overseer squarely in the spotlight, but another critical player in U.S. nuclear safety is the NRC's Office of the Inspector General, an independent agency that serves as watchdog to the watchdog.

Now, Mulley and one other former OIG employee have come forth with allegations that the inspector general's office buried the critical Byron report and dropped an investigation into whether the NRC is relying on outdated methods to predict damage from an aircraft crashing into a plant.

The inspector general's office, they assert, has shied away from challenging the NRC at exactly the wrong time, with many of the country's 104 nuclear power plants aging beyond their 40-year design life and with reactor meltdowns at Fukushima rewriting the definition of a catastrophic accident.

"We're in the nuclear power business. It's not a trivial business; it's public health and safety," said Mulley, who won the agency's top awards and reviewed nearly every major investigation the office conducted before he retired as the chief investigator three years ago.

"We have to have somebody that's going to look over the NRC's shoulder and make sure they were fulfilling their obligations," he said.

Inspector General Hubert T. Bell declined to comment, but Joseph McMillan, the assistant inspector general for investigations, said the office has continued to vigorously pursue cases. He confirmed that the aircraft crash case has been closed but said it was proper. Regarding the Byron case, McMillan acknowledged disagreements but said: "I stand by the work we have done."

The U.S. nuclear industry can point to an enviable safety record -- no member of the public has ever been injured by an accident at a plant. Nonetheless, critics point to issues like the NRC's drawn-out effort to enforce fire rules as evidence that the five-member commission and the agency it runs are too close to the industry.

The inspector general's office has traditionally filled a key oversight role, conducting dozens of investigations that have changed how the NRC regulates nuclear waste, fire protection and security, among other things. Its regular reports to Congress cover waste, fraud and agency performance.

Many federal agencies have similar independent offices to ferret out wrongdoing and improve efficiency. The NRC's was established in 1989 and has been led for the past 15 years by Bell, who was appointed by President Clinton after nearly three decades in the Secret Service.

'Everything Seems to Die'

In the office's history, Mulley has left a big mark.

For years, he documented how the NRC dropped the ball on the handling of nuclear fuel and security in nuclear plants. His reports on defective fire barriers led to congressional hearings and ultimately to a complete overhaul of the agency's fire protection regulations.

He retired in 2008 as a senior-level assistant for investigations but continued work as an OIG consultant for two more years. Before he retired, Bell and a deputy wrote that Mulley was "so thorough and knowledgeable of all aspects of investigations, that even NRC management recognizes the value added to having Mr. Mulley's expertise on all cases."

Mulley is not alone in his concerns about the inspector general's office. Another former employee told ProPublica that the office has become reluctant to probe anything that could become controversial or raise difficult questions for the NRC.

"They don't want to do anything," said the ex-employee, who left out of dissatisfaction with the direction of the office and asked not to be named to protect his current job. "Everything just seems to die."

The former employee told ProPublica that the OIG's office had dropped an inquiry into whether the NRC could accurately predict the damage to a plant from an airplane crash, and Mulley confirmed his account, saying the office received a tip in 2007 that the NRC was using an outdated method.

Because a wrong prediction could lead to insufficient protection for the plants, the inspector general's office opened an investigation, Mulley said. "We went to several experts who said that thing is antiquated, you can't use it," he said.

Mulley said that the NRC's experts insisted that their method was accurate. He said the aim of the investigation was not to prove that the NRC experts were wrong but to show there was a dispute and question whether the NRC should update its predictions.

"In my mind, the OIG was not going to resolve it," he said. "It raised a valid question."

The 2001 terrorist attacks drew attention to the potential hazard of an aircraft crash for nuclear plants, and afterward the NRC and nuclear industry examined whether new precautions were needed.

The main industry trade group, the Nuclear Energy Institute, commissioned studies that showed U.S. plants could sustain a direct hit from a modern airliner without any radiation release.

Following 9/11, the NRC adopted a rule requiring nuclear operators to take steps to minimize possible damage from major natural disasters or an aircraft crash. Two years ago, the commission required new licensees to assess whether their reactors could withstand an airliner crash.

Eliot Brenner, an NRC spokesman, said the agency's method of evaluating the risk to plants has been thoroughly checked and relies on "realistic threat parameters."

McMillan said that OIG completed its investigation into the crash prediction issue and that the case was "closed to the file," meaning that no report was issued.

The decision to forgo a report usually means that the inspector general found no public safety concerns. McMillan declined to comment on the report or to describe any conclusions. He said it was available only through a Freedom of Information Act request, which ProPublica filed today.

The Byron Plant's Rusty Pipe

Mulley spent more than a year investigating why the pipe blew out at the Byron plant.

On Oct. 19, 2007, a worker using a wire brush to clean a thick coating of rust from the massive steel pipe ripped completely through the metal. Water shot out, triggering a 12-day shutdown of the plant's two reactors located outside Rockford, Ill.

The 24-inch pipe was part of the plant's Essential Service Water System, a network of eight huge pipes that carries water to cool emergency equipment. During an accident, it can be critical because it protects the generators and pumps that keep the reactor from overheating.

"It's a safety-related system," Mulley said. "If it doesn't operate, you can't operate the plant."

After the pipe ruptured, the NRC assigned a special inspection team to find out whether Exelon could have prevented it. Mulley put together a four-person team to start a parallel investigation into whether the NRC inspectors should have caught the problem beforehand.

His team interviewed workers and NRC inspectors assigned to the Byron plant since the early 1990s. They concentrated heavily on the inspectors' actions in 2007, when Byron engineers began scrutinizing pipe sections, called risers, that were partly buried in concrete in a below-ground vault.

Plant engineers performed ultrasonic tests on the thickness of the risers. Originally, the pipe walls were three-eighths of an inch thick, but over the span of three tests, engineers stepped the acceptable thickness down to three-hundredths of an inch -- equivalent to seven sheets of paper.

Mulley's team found that the NRC's on-site inspectors had not checked the Byron engineers' work even though repeated drops in safety margin should have been a red flag. Corrosion in Byron's essential water system had been discussed in plant meetings, and because testing the risers required repeated use of a crane to gain access, inspectors should have suspected something.

"The NRC is supposed to -- if they're overseeing this thing -- take a look at it and say, 'Oh, wait a minute, what's going on?'" Mulley said. "But obviously, they didn't look at that one."

Mulley found that NRC's on-site inspectors had repeated opportunities to check the pipes over the years but had not done so. In interviews, the inspectors told Mulley's investigators that they had been busy with other work. Although inspectors had preformed a required number of equipment checks, Mulley's report found that their inability to set priorities was a weakness in the inspection program.

The NRC, it turns out, had received a warning about a similar pipe break at the Vendellos nuclear plant in Spain, Mulley's team discovered. Peter B. Lyons, then an NRC commissioner, had even mentioned the Vendellos break in a speech, saying the agency was on top of the problem. But the word was never sent to NRC inspectors in the field, Mulley found.

"I don't think anybody up there was purposely saying, 'Hey, this is not so important,'" Mulley said of the Vendellos information. "I think they knew it was important. I think they intended to. I don't think anybody followed up on it, and then it falls into the cracks."

Report Revised, Kept From Public

Because the Byron incident touched broadly on NRC inspection policies, Mulley opened his case as an Event Inquiry -- a report normally intended for release to Congress and the public. He stayed on after retirement to complete it, submitting it in 2009 with some tough conclusions.

The NRC "provided little meaningful regulatory oversight of corrosion of piping in the Byron essential service water system, one of Byron's most risk significant systems," his version states. Moreover, the NRC "did not take full advantage of lessons learned" from Vendellos.

Mulley said no one raised questions.

"The report languished for a year," he said. "Nobody ever got back to me once to let me know, although I emailed them asking what's going on, what's happening with this thing."

Then, in September 2010, the inspector general's office issued a new version. Mulley's draft had been thoroughly rewritten, and although the facts were similar, the conclusions were not.

The report said NRC oversight "was not successful" and that guidance for inspectors "was not specific enough," but pointedly blamed Exelon for the inspectors' failings.

"Although the (NRC) resident inspectors carried out routine oversight responsibilities in accordance with agency requirements, the licensee's failure to analyze a problem correctly resulted in the resident inspector's lack of awareness of a significant problem," it states.

By contrast, Mulley's version squarely faults NRC inspectors and procedures.

"From 2000 to 2007, the NRC did not conduct any documented inspection activity of essential service water piping," it states, while inspectors "provided no regulatory review … to support the licensee's lowering of the acceptable minimum wall thickness" in the piping.

The revised report did not mention Vendellos or the NRC's failure to inform inspectors about it. And instead of being issued publicly, the report was classified for internal use only.

"I was amazed," Mulley said. "This had never happened before in all my years."

Mulley said the official report left out systemic problems his team uncovered and was not published so that shortcomings in NRC oversight would be hidden from the public and Congress.

"I think changes that could have been made, pressure that could have been applied to improve the process, improve our oversight, are not going to be done," Mulley said.

'We Stand by the Report'

Brenner, the NRC spokesman, said the commission has upgraded procedures as a result of its own review of the Byron incident. In particular, he said inspectors were told to prioritize inspections of areas that had limited access and of equipment that repeatedly degraded, like the pipes at Byron.

McMillan declined to answer any specific questions about the Byron report because the matter has been referred to the Council of the Inspectors General on Integrity and Efficiency, which has the authority to investigate allegations of wrongdoing against inspectors general.

He said he believed the Byron case was handled appropriately. "We can have disagreements over how the reports are handled," he said, "but at the end of the day, we stand by the report."

A spokesman for the council's integrity committee said he could not comment. Marshall Murphy, an Exelon spokesman, also declined to comment. The company previously has said it improved procedures after the pipe rupture at Byron.

The significance of a strong, independent inspector general is not lost on the NRC, which is struggling with how to respond to the Fukushima accident after a special agency task force called for a potentially far-reaching reworking of regulations covering catastrophic events.

Commission Chairman Gregory Jaczko, who has come under fire recently for pushing too fast on reforms, reflected on the inspector general's role in a statement last month.

The office, Jaczko said, "plays an important role in enabling the American people to continue to have confidence that my focus as chairman -- and the entire agency's focus -- is on effectively carrying out the NRC's vital safety mission."

Mulley said that mission is too vital for him to remain silent.

"I am coming forward because I spent my entire life, most of my professional life, doing this," he said. "We get the power to write these reports, we get the power to talk to you. We've got the power to go to (Capitol) Hill, at least keep it in line a little bit as much as we can.

"We can't be every place but at least try to keep them in line, and I think it's vital."