A string of mechanical and human errors that led to the ongoing massive oil leak in the Gulf of Mexico began when the cement poured into the bottom of the well failed to do its job, BP PLC investigators said yesterday.

The cement was supposed to seal off the reservoir of crude and natural gas that had been discovered at the Macondo well, allowing workers to begin the process of shifting from exploration drilling to production. But even once the cement failed, the oil and gas still would not have been able to flood the wellbore if it were not for a series of other missteps and failures, BP said.

The London-based oil giant made public for the first time yesterday details from its ongoing investigation into the causes of the April 20 explosion on the Deepwater Horizon rig that killed 11 workers and started the leak that continues to spew untold amounts of crude into Gulf waters more than a month later.

The nearly 50-page PowerPoint presentation shown to reporters yesterday afternoon provides the most detailed public account to date of what BP says it believes were the causes of the explosion and subsequent oil leak.

In addition to the faulty cement, investigators believe at least one of two secondary barriers also failed to prevent the oil and gas from entering the wellbore, as did at least two other emergency shutoff mechanisms that were designed to seal off the well and prevent the escape of the massive amounts of oil that are now reaching coastlines along the Gulf Coast.

BP also said workers apparently failed to recognize the leaking natural gas that ultimately ignited the rig, as well as several other signs that efforts to seal the well were not going as planned.

While providing an extensive technical briefing -- which lasted more than two-and-half hours -- of an unfinished investigation so soon after such a major accident is typically rare, BP officials did withhold some key parts of their ongoing work.

BP officials refused to name names regarding which companies were responsible for the building and installation of the equipment they believe was faulty -- a major issue that is sure to prove a point of contention as lawmakers and the Obama administration conduct their own investigations and look to assign blame, not to mention civil and possibly criminal penalties.

"We're not going to get into those specifics," said Dave Nagel, vice president of BP America. "We're just here trying to analyze this in terms of what happened and the mechanics of it, rather than who made it, who installed it -- we're just not going to get into that."

BP leased the rig from Transocean Ltd. and contracted Halliburton Co. to do the cementing and other tasks. Earlier this month President Obama slammed the three companies for "falling over each other to point the finger of blame at somebody else" during their first trips to Capitol Hill.

BP also refused to release the names and title of three members of the investigative team who helped brief the media, citing the ongoing investigation. Likewise, the company declined to provide electronic copies of the presentation to prevent it from being posted online.

Explosion and secondary barriers

Investigators said it was still unclear how the gas ultimately moved from the wellbore to the rig but that once it began to escape from the well in the quantities it did an explosion became nearly unavoidable.

"That was a pretty large plume of gas," said Mark Bly, BP's head of safety and operations. "It was large enough that there was a high likelihood that it would have ignited."

The influx of gas, along with the failure of the rig crew to notice it, were among the most critical moments leading up to the two explosions that ultimately sunk the Deepwater rig. According to eyewitness accounts, gas detector alarms were going off shortly before the explosions, which occurred seconds apart at 9:49 p.m.

"The hydrocarbon got out, the gas plume got out and found an ignition source," Nagel said.

While the cement served as the primary barrier between the well and the hydrocarbon reservoir more than 3 miles below the water's surface, it was far from alone in being responsible for keeping the oil and gas from escaping.

Once the oil and gas penetrated the cement it would then have had to overcome one of two secondary barriers to enter the well hole, the BP investigators said.

One possibility was that it passed through the so-called float collar, positioned at the bottom of the wellbore. The second would have been for the oil and gas to travel up the outside of the well to roughly the seafloor, where it then would have had to pass through a device known as a seal assembly and back down into the wellbore. It was not clear which companies built the two devices.

"At this juncture, the investigation team does not know which one of those happened," Bly said. "I think we may not ever figure out which of the two it was."

After the explosion, two emergency devices -- the blowout preventer, or BOP, and the "dead-man" function -- also failed to do their jobs. According to the presentation there was no evidence either activated properly.

Missed signals and dispute

During the planned operation to seal the well, rig workers performed a series of tests to ensure that it was secure.

While the first several tests proved successful, toward the end of the process there were a number of results that should have caused rig workers to be concerned, BP said.

"The integrity testing was not understood by the people on the rig," Bly said. "They thought they had achieved isolation of the reservoir."

One of the clearest signs was the fact that during one test, the drill pipe and another smaller pipe had different pressures when they should have been equal.

"This is, we believe, where a mistake was made in the process," said one of the three unidentified investigators. "It demonstrates that there was something not right at the well at that time. For this test to be successful, it has to stay zero and zero on both lines."

The investigators said that at that point, according to witnesses, a discussion between unidentified rig supervisors took place but that it was not clear why they decided to proceed. "We don't know why, we're in the same camp you all are," said one of the BP investigators.

While the BP officials did not name the supervisors, or their employers, during the Washington briefing, details of the dispute emerged in Louisiana at a hearing led by the U.S. Coast Guard and the Interior Department's Minerals Management Service, which jointly regulate offshore drilling.

According to media reports, Transocean's chief mechanic on the rig described a "skirmish" between a Transocean official and one from BP during an 11 a.m. meeting on the day of the explosion.

When asked whose decision it would have been to proceed despite the test results, BP investigators in Washington declined to say. "I'd rather not offer an opinion," said one.

Another signal that should have tipped off rig workers, BP said, was that two hours before the explosion the system collected three times the amount of fluid that it would normally, another sign their interpretations of test results were wrong.

Meanwhile, one of BP's men on the rig when it exploded, Robert Kaluza, will not testify today as planned before the investigative panel in Louisiana, asserting his Fifth Amendment right not to incriminate himself, the New Orleans Times-Picayune reported.

Donald Vidrine of BP also was scheduled to testify, but he has withdrawn, citing illness, the Coast Guard said.

Questions remain

BP officials stressed that investigators had still not collected all of the information or interviewed all of the witnesses and that new data could change many of their current findings.

"This is not an interim report by any means," said Nagel, referring to the preliminary nature of the findings. "We're not going to try to draw conclusions."

BP said the investigation began shortly after the rig exploded on April 20 and that there are now roughly 70 personnel on the team. Much of the current findings are based on real-time data transmitted to BP's Houston office from Halliburton and other contractors operating on the rig.

Nagel said that in addition to why the rig workers missed numerous signals, investigators were also trying to find out why attempts to close valves on the BOP in the final minutes before the explosion were unsuccessful.

He said investigators were also checking whether a decision by BP officials to run less than a third of the planned 21 tests to ensure the drill pipe was centered in the well hole played a role in the accident. If the pipe had been off center, it could have caused flaws in the surrounding cement.