As engineers bore deeper into the seafloor toward the source of the oil still spewing into the Gulf of Mexico, BP PLC is growing more confident that the relief well it expects to complete in August will succeed where all previous efforts to contain or kill the gusher have failed.

But what if it doesn't work?

At the very least, oil would continue to spill while workers try something else.

That proposition would surely bring more misery for the people who live, work and play along the shores from Louisiana to Florida.

And consider this: Chief Executive Tony Hayward said in June that the reservoir of oil is believed to hold about 2.1 billion gallons of oil. If the problem was never fixed, it could mean another two years of oil spilling based on the current flow rate until the reservoir is drained.

BP says the first relief well is on target to be completed by early August. A second relief well, which could be completed a few weeks later, is viewed as a backup if the first one doesn't work.

But efforts to contain other major oil spills haven't always gone according to plan.

The 1979 Ixtoc oil spill, the Gulf's worst oil spill before it was eclipsed by BP's disaster, wasn't contained until three months after the first of two relief wells was completed. By then, 140 million gallons of oil had spilled in the 10 months it took Pemex, Mexico's state-owned oil company, to stop the leak.

That's why BP is developing "backups for the backups." But the British company is sparse on details, and even the ideas it is floating can't guarantee the blown-out well that has already pumped up to 160 million gallons of oil into the sea over 2 1/2 months won't keep flowing into the fall — or perhaps even beyond.

So, the Gulf region is left to hold its collective breath as BP puts much of its effort into the relief well just as Mother Nature could unleash a blistering hurricane at any moment.

"The relief well itself is not a slam dunk," said Gene Beck, a petroleum engineering professor at Texas A&M University.

Kent Wells, a BP senior vice president, said other options include trying to reconfigure the existing containment cap to collect more of the spewing oil or tying it into another production platform on the surface. However, Wells has been mum on a game plan and he said no decisions have been made on the alternate platform idea.

BP declined repeated requests from The Associated Press over several days to make Wells available to elaborate or for a spokesman to comment further.

As to the hurricane concern, Wells said only that the rigs drilling the first relief well and the backup relief well are designed to operate in everything except a tropical storm or hurricane. If engineers had to disconnect and evacuate the area, drilling could be offline for 14 days, during which time an estimated 2.5 million gallons of oil would flow into the Gulf unabated each day.

History is on BP's side, but the depth of the seafloor isn't.

Engineers and oil industry experts familiar with or involved in previous relief well missions at sea say that if the heavy mud BP plans to pump into the existing well from underneath at its source doesn't stop the flow altogether, it should at least reduce the pressure that is forcing oil so fast into the sea.

Carlos Osornio, a Mexican engineer in charge of Pemex's deepwater drilling operations during the Ixtoc crisis, said BP may ultimately find that both relief wells are needed to contain the gusher.

"One relief well may not be enough to contain the high volume (of oil flow), but two will work for sure," he said.

A reduction in pressure could give BP the option of putting a new blowout preventer on top of the one that was damaged in the April 20 explosion. That was a containment option BP considered early on, but hasn't tried because of the risk posed by the amount of pressure from the seafloor.

A new blowout preventer isn't foolproof either.

"It's very unpredictable because the current condition of the well down there is unknown," said Satish Nagarajaiah, a Rice University engineering professor who focuses on offshore structures.

BP engineers are using tools and running tests that tell them where they need to go. Drilling down parallel to the gushing well before cutting in sideways makes that data more accurate than it would have been if they were approaching the well horizontally, said Donald Van Nieuwenhuise, a University of Houston geology professor who has been a lead geologist on several offshore drilling projects.

"They're not looking for a needle in the haystack anymore," he said. "Now they're just trying to figure out where they want to pick that needle up."

Bruce Bullock, director of the Maguire Energy Institute at Southern Methodist University, said he is "somewhat suspect" that the relief well will hit its target on the first try.

"You're going 18,000 feet to hit a dinner plate. My guess is two or three times is more of a likelihood," he said.

Osornio, the former Pemex engineer who is now a deep drilling consultant, said there is no reason BP wouldn't be successful the first try.

"Today's tools provide specific locations in real time as they drill, something we didn't have during Ixtoc," he said.

Still, there's potential peril if BP misses its target and decides to drill deeper directly into the oil producing formation.

Engineers tried that approach and were successful in killing several out of control wells in 1970 during the Bay Marchand fire off Louisiana.

But George Hirasaki, a Rice University professor in chemical and biomolecular engineering who was involved in the Bay Marchand oil containment effort for Shell, said engineers have to be very careful when drilling into any formation that has hydrocarbons, which poses the risk of the same type of explosion that destroyed the rig.

Bullock said there have been past successes with relief wells on land and in shallower waters, but no relief well is risk-free.

Beck said he expects the drillers to hit their mark on the first try but wouldn't be surprised if it took two or three attempts. Beck puts the odds at 80 percent that the relief well will in short order kill the gushing well.

"There haven't been a significant number of deepwater blowouts before," he said. "To a certain extent, we're in an unproven area here, as well."


Kunzelman reported from New Orleans. AP writer Peter Prengaman contributed to this report.

Mochila insert follows...