WASHINGTON – The saga of BP’s runaway Deepwater Horizon well, already entering its third month, has entered a crucial phase that will determine whether the Gulf of Mexico gusher ends in mid-August or persists, perhaps for months.

Unlike the previous public drama, this act will unfold miles below the seabed, as drill technicians begin delicately maneuvering a relief well that they hope will pierce and cap the gushing oil well.

Additional Photos

The drilling rig Discoverer Enterprise recovers oil from the leaking Deepwater Horizon site in the Gulf of Mexico earlier this month. BP is deploying deep-water sensors to better determine the amount of oil leaking from the uncapped well, at the same time that work continues on drilling two relief wells. The Associated Press/The Sun Herald

This week, BP began using sensitive electronic equipment to detect differences in the rock’s electromagnetic field in an effort to pinpoint the metal pipes inside the wellbore. Based on what they find, they’ll make adjustments every few hundred feet in an effort to intercept those pipes and kill the gusher by pumping it full of tons of heavy drilling mud and then concrete.

The stakes riding on those adjustments are enormous, and the chance of failure, at least on the first try, is huge.

“The engineers will tell you that they have a 95 percent chance of success” in killing a runaway gusher with a relief well, said Bruce Bullock, the director of the Maguire Energy Institute at Southern Methodist University in Dallas. “But that depends on how you define success. It’s quite unlikely they’ll hit it on the first stab.”

“They’re aiming at a salad plate thousands of feet down,” Bullock said: a 7-inch pipe buried in concrete, 12,000 feet below the seafloor.

Every time a relief well misses, its crew must back up the drill bit and try again. Last year, a relief well aimed at capping a blowout in the Timor Sea off Australia missed its target four times before connecting. Each new effort took an average of another week of drilling, for a total delay of 27 days after the drillers began closing in.

A similar delay at the Deepwater Horizon site would mean as much as 1.62 million barrels more crude dumped into the Gulf — more than 68 million gallons — if the latest government estimates of the flow are accurate.

Many obstacles lie in the way of the relief well, not least of which are the same vagaries of subsurface strata and gas pockets that put the Deepwater Horizon 43 days behind schedule before bad decisions and equipment failures sent it to the bottom of the Gulf.

On Wednesday, drillers were reminded how quickly success can give way to desperation.

Only a day earlier, a jury-rigged containment system had captured more crude than ever before, 27,090 barrels. Then a robot vehicle nudged the “top hat” containment dome, closing a vent and sending what workers feared was flammable natural gas up a line to a drill ship on the surface.

Mindful that an unexpected surge of methane gas had doomed the Deepwater Horizon, they quickly yanked the containment dome away, leaving the well to gush, unimpeded, until the dome was put back in place 10 hours later.

“Things have gone well, down to this stage,” Kent Wells, BP’s senior vice president of exploration and production, said a week ago about the relief well drilling. “Of course, that doesn’t always mean things will continue to go well.”

His caution is echoed by Coast Guard Adm. Thad Allen, the Obama administration’s point man on the Deepwater Horizon catastrophe.

“They are slightly ahead of schedule,” Allen said of the relief wells during one of his daily briefings this week, “but I’m not coming off the second week in August date because we know, you know, things can happen.”

Uncertainty awaits even once the drillers pierce the Deepwater Horizon pipe. That’s when they’ll try to plug the well with drilling mud and concrete. The idea is that the heavy mud will counteract the pressure that’s forcing the oil up the well pipe.

However, such plans have failed before, according to studies of previous relief well attempts, either because the mud didn’t weigh enough or the drillers didn’t have enough of it.

Officials also acknowledge that the plan could be defeated if the well was seriously damaged by the explosion of methane gas that sank the Deepwater Horizon rig and took a mile of pipe down with it.

Congressional investigators have accused BP of causing the blowout by cutting corners on the well’s design, not using enough spacers to hold the well pipe in the center of the bore in order to save money and selecting a single-pipe system for the final leg of the well, instead of the usual one-pipe-inside-another approach.

Those choices also could undermine efforts to kill the well, by allowing pathways for the drilling mud and crude oil to seep into the surrounding rock instead of filling the well pipe.

Adding to the uncertainty of the relief well effort is the fact that, like so much of the Deepwater Horizon containment effort, there’s no good precedent for comparison.

The last great gusher in the Gulf, the Ixtoc 1 blowout off the coast of Mexico in 1979, ran wild for 10 months before relief wells found it and killed it. That, however, was before the latest electromagnetic technology was developed in the mid-1980s, when petroleum engineers realized that they could increase the accuracy of their ranging efforts by running electricity through the runaway well’s metal pipe.

The blowout off Australia last August is a poor guide. That well was in 250 feet of water, not the 5,067 feet where the Deepwater Horizon was drilling, and the oil reservoir was less than half as deep as the Deepwater Horizon’s Moncado field in the Gulf.

At those depths, adjusting the drill bit to make a second, third or fourth approach to the well is likely to take more time.

“The operations at the well itself are so complicated,” SMU’s Bullock said. “The mechanics take a while to do.”

BP’s Wells said last week that the wellbore now lies tantalizingly close, just 200 feet laterally from the first relief well. That puts it within the limits of current electromagnetic ranging capabilities.

At 10,677 feet below the seafloor, however, the first relief well is still at least 1,000 vertical feet above where BP’s engineers would like to enter the Deepwater Horizon wellbore. While electromagnetic ranging helps, it can be off by as much as 10 percent, according to a study by John W. Wright, one of the world’s foremost experts on relief well drilling.

So for the next several weeks, drillers will take a path down parallel to the Deepwater Horizon’s wellbore while trying to inch closer, guided by electromagnetic pulses in a process that in Wells’ descriptions sounds equal parts science and trial and error.

“We’ll drill a little bit, then we sort of use our tools to sort of test where we think it is,” he said. “We’ll pinpoint a place, then we’ll drill another couple hundred feet towards it, test again and see if we’ve, you know, made the right progress.”

The drama isn’t lost on the hundreds of men who are drilling the wells at the Deepwater Horizon site.

“I’m glad to be part of the operation to try to stop this well,” Wendell Guidry, a senior tool pusher aboard the Development Driller II, said last weekend. The spill, he said, is “not only affecting me, but my family, my kids, later on down the line.”

Related Stories