The freighter slammed into the wharf at 2:11 p.m., collapsing huge chunks of the structures sitting above it. Amid the screams and the blare of the horn came the terrifying sounds of crumbling concrete and screeching metal. The freighter, loaded up with corn, then bounced off the wharf and headed toward the riverboat casino. A few of the Flamingo’s eight hundred gamblers, after instantly calculating their odds, jumped overboard. Mercifully, the freighter came to a shrieking stop—seventy feet shy of the Flamingo. “It felt just like an earthquake49,” one survivor told the Times-Picayune. “It really just seemed like it was something that happened on TV.”
The crash of the Bright Field, operated by a Chinese crew sailing under a Liberian flag, ended up injuring sixty-seven people and taking out fifteen stores and restaurants, as well as parts of the nearby parking garage and hotel. In all, it caused $20 million in damage. Investigators said it was a miracle no one had been killed. And they attributed that good fortune in part to the quick instincts of the freighter’s American pilot, who had wasted no time in ordering that the Bright Field’s anchors be dropped to try to slow the beast’s slide into the shopping mall.
Those were precisely the kinds of quick instincts that gave the designers of the Deer Island Outfall Tunnel nightmares.
There was more than ocean sitting above the Deer Island tunnel. The ribbon of Massachusetts Bay directly overhead just happened to be a busy shipping lane from the Atlantic into Boston Harbor. The decision to site the tunnel under that lane had been a fragile compromise following contentious negotiations, so there would be no going back on it. Yet the site path left the tunnel’s designers fearful that a ship’s anchor might dislodge one of those fifty-five domed diffuser heads sitting on the ocean floor. It was an extremely remote risk, since even if an anchor made contact, the domes were designed to withstand a direct hit. Still, the designers worried about a freak accident in which a captain, after losing control of his ship in a storm, began dragging his anchor in an emergency attempt to stay afloat. Although the tunnel drawings had been finalized long before that freighter took out the mall in New Orleans, the Bright Field accident represented the kind of worst-case scenario the designers had in mind.
Hydraulic studies predicted50 that if an anchor ever did succeed in ripping off a diffuser head during tunnel construction, the subterranean sandhogs would find themselves thrust into a disaster-movie scene even scarier than the one featuring all those Big Easy shoppers. Once the diffuser cap was breached, ocean water would gush down that riser and flood the empty tunnel at a crushing rate of 115,000 gallons per minute. It would resemble a flash flood51, with a leading wall of water as wide as the tunnel and as high as five feet, moving at a rate of nearly four feet per second.
Designers shuddered at the prospect of an accident like this endangering both the expensive structure and the scores of workers who were building it. The picture of fifty or a hundred sandhogs racing for their lives against this violent wave of seawater was a terrifying one. The lokey-pulled tram would be of no use, of course, so the workers would be forced to try to either outrun or outswim the seawater. Even if they were able to access emergency rafts, many would likely not make it out alive. To guard against this disaster movie ever playing out in real life, the tunnel’s design firm,
Parsons Brinckerhoff (PB), had insisted that contractor Kiewit install a jumbo safety plug at the bottom of each of the fifty-five risers.
In the fall of 1996, just a couple of months before the New Orleans accident, the sandhogs had managed, finally, to get the TBM to its 9.5-mile finish line. To reach the end, man and machine had excavated an astonishing 2.4 million tons of rock52. It was impossible to drive the gargantuan TBM in reverse all the way back to Deer Island, because the tunnel’s thick concrete lining had made it narrower than the machine itself. Instead, Kiewit instructed the sandhogs to use the TBM to keep boring an additional hundred feet to make a grave. Then, like something out of a mob hit, the $15 million bullet-shaped behemoth was buried in concrete.
This milestone came a full year after the federal judge’s original deadline for the completion of the entire tunnel, and huge challenges remained. One of the most formidable involved turning the final stretch of the tunnel, which was the standard diameter of twenty-four feet, into something much narrower. Because gravity would be the only force powering the millions of gallons of sewer water along the miles of tunnel and up those fifty-five risers, the wastewater would have to be kept moving with enough velocity that any solids in it remained in suspension. Otherwise, the solids would settle to the bottom of the tunnel and eventually clog it. This same principle had guided the design of sewers going all the way back to Roman times. A good flow would also ensure that the sewer water climbed up the risers and exited out to sea with enough force to prevent an intrusion of ocean water from coming in from the other direction—a real risk since the salt in the ocean water made it slightly heavier than the treated sewer water. Those velocity demands meant the tunnel had to get progressively narrow along its last mile, as it connected with each of the fifty-five risers.
Kiewit chose a three-pronged approach53 to reduce the area inside the tunnel. First, workers built a long ramp, beginning near the start of the final mile, that would gradually reduce the tunnel’s top-to-bottom space. Then, about halfway into that last mile, they installed a series of seven-foot-tall Jersey barriers—like the ones that line highway medians—to reduce the tunnel’s side-to-side space. The farther east the tunnel went, the closer together those Jersey barriers were positioned. Finally, for the last six hundred feet or so of the tunnel, they laid in several precast pipes of decreasing diameters and then backfilled around them with concrete. The last of these precast pipes—which would house the connections to risers number 3, 2, and 1—was just five feet in diameter.
Meanwhile, sandhogs needed to finish hooking up the tunnel to all the risers. They had to repeat, fifty-four times, the same task that Mad Man and the others had completed at riser number 55, when thousands of gallons of Gatorade-green water had come gushing out. At each hookup, once the probe hole had been drilled and the riser had been drained of its green water, a new phase of excavation would begin. The sandhogs essentially had to dig fifty-five small side tunnels. Doing this allowed them to install concrete-encased elbow pieces that connected each of the thirty-inch-wide vertical risers with a horizontal riser pipe of the same width that would shoot off from the side of the main tunnel. Those elbows would create54 the clear path through which treated sewer water would eventually flow from the Deer Island plant, through the tunnel, up through a riser, and out to the ocean floor. The elbows had another important function. Pre-installed in each one was a sixty-five-pound domed safety plug that resembled a salad bowl from some industrial kitchen. Those plugs were designed to help protect the sandhogs from the ocean above.
By 1997 all fifty-five of these elbows and their safety plugs were in place. Only then did the moment of reckoning seem to begin for Kiewit: When everything else is done, how in the world are we going to get these plugs out? There was no good answer to this question because of an unusual provision buried in Kiewit’s tunnel contract, a provision that would make the task of removing those plugs infinitely harder.
49 (Chapter 3, note 3)
“It felt just like an earthquake”: Chris Gray, “Workers, Shoppers Run for Safety in Chaos,”
50 (Chapter 3, note 4)
Hydraulic studies predicted: Eldon Abbott, memo to Paul Zick and Dave Corkum, “Outfall Tunnel—Tunnel Flooding Due to Loss of Diffuser Head,” December 15, 1997; Dave Sinsheimer, memo to Ken Stinson and Ron Minarcini, “Deer Island Hydraulic Study,” December 19, 1997.
51 (Chapter 3, note 5)
It would resemble a flash flood: One hydraulic study estimated that the wave would race through the entire 9.5 miles of the tunnel and make it to the base of the shaft in about three and a half hours. It would likely take workers a lot longer than that to get out of the tunnel since, during a flood, the lokey would be of no use. Within another eight hours, the height of the wave would reach the top of the tunnel.
52 (Chapter 3, note 6)
2.4 million tons of rock: Eric Niiler, “MWRA Finishes Drilling Sewage Outfall Tunnel,”
53 (Chapter 3, note 7)
Kiewit chose a three-pronged approach: Tom Corry (Kiewit executive and project “sponsor”) deposition in civil lawsuits filed by divers and families, June 2000, 72-79.
54 (Chapter 3, note 8)
Those elbows would create: These side tunnels, which the sandhogs dug mostly with their jacklegs, ranged in length from nine feet to thirty-five feet, depending on how far that particular riser was located from the main tunnel. After digging their way out to each vertical riser, the sandhogs would attach the prefabricated, concrete-encased elbow piece to the bottom of that pipe. Once the elbow and the horizontal connector pipe were in place inside each of these side tunnels, crews would pump in concrete to backfill around the connector and riser pipes. With that, the width of each of the fifty-five openings, from the main tunnel all the way up to the ocean floor, was no wider than thirty inches.