9. Tuesday
When the school bus ferrying the divers arrived on the island at seven o’clock on Tuesday morning, Harald was on the phone with Tony Drybanski. The salesman-president of the mixer’s local distributor was responding to Harald’s message from the night before. He sensed Harald was agitated and impatient for an answer, but that was nothing new. “Every time Harald called178, he spoke to you like a debt collector,” Drybanksi would later say. “So he always had this sense of urgency in his voice. He was just all over me, all the time.” To address the MAP Mix 9000’s flow problems, Drybanski advised Harald to bump up the pressure on the machine. That couldn’t be done simply by adjusting one of the two dials on the front of the device. Instead, he said, Harald would have to unscrew the mixer’s stainless-steel cover and manually adjust the internal gauge that governed the pressure of the oxygen and nitrogen coming into the machine. The first page of the owner’s manual warned that the “MAP Mix 9000 should be opened by authorised personnel only,” and Harald clearly didn’t fall into that category. Still, based on his previous discussions with people at the factory back in Denmark, Drybanksi explained to him that this inlet pressure gauge could be safely turned up. The mixer had left the factory with a default setting of 4.3 bar of pressure, but Drybanski said the gauge could be increased to 7 bar, or the equivalent of about 100 pounds per square inch (PSI179). Rather than max it out, they settled on 6.6.
The goal was to get more juice coming out of the mixer. But there was apparently no discussion between Harald and Drybanski about how raising the pressure of the gas flowing into the mixer might affect the inner workings of the blending process. After all, the magic that powered the pneumatic MAP Mix 9000 was not electricity or gasoline but pressure—or more precisely, a differential in pressure, based on the way gas naturally moves from high pressure to low. To ensure that the machine’s valves and diaphragms functioned properly, the inlet pressure coming from the oxygen and nitrogen tanks had to be at least 1 bar higher than the outlet pressure of the blended gas going out into the divers’ umbilicals180.
Harald took the man cage down to the base of the shaft. Enlisting DJ’s help, they popped off the lid and made the adjustment. Harald then ran a test by turning on the liquid oxygen and liquid nitrogen and checking the flow meter on the manifold. The test showed181 that the increase in pressure created an increase in the flow rate—the gas came out faster and more forcefully. The adjustment had worked. Harald told the divers that the simple maneuver should address Hoss’s complaints about the system.
Because his lungs were still hurting him from his facemask problems on Monday, Hoss would be staying out of the tunnel today. The Tuesday team consisted of Tim and DJ in the Humvee, and Billy, Mars, and Riggs on the excursion crew.
It was just after nine-thirty a.m182., following the lengthy preparation and checklist period, when the divers piled into the Humvees and began their drive. About three hours later, they arrived at riser number 12, and Tim called Harald to tell them they were setting up. After the excursion crew started moving east with the boat, DJ and Tim settled into their places in the Humvee wagon.
Because DJ had been out of the tunnel on Monday, this was his first experience working with the peculiar breathing system. With Tim behind the wheel, DJ sat in the back seat, next to the MAP Mix 9000. Like the trio on the excursion crew, DJ and Tim wore their facemasks, which supplied them with the blended air coming out of the mixer. It was dark inside the Humvee. The engine wasn’t running, but they drew power from the battery to keep the vehicle’s dim dome lights on. When Tim used the comm wire to talk with Mars, who was once again serving as the foreman of the excursion crew, DJ could hear both ends of the conversation. But when Tim was talking to Harald on the mine phone, DJ could hear only Tim.
As Billy had done the day before, DJ monitored the breathing system serving all five divers. When Harald had set the mix dial on the front of the machine to 22, he was telling the mixer to blend oxygen and nitrogen in such a way that it produced breathing air containing 22 percent oxygen. But that was effectively just a request. There was no gauge or monitor on the mixer183 to confirm that the actual blended air was meeting the desired amount of oxygen. DJ’s job was to use the Drager Multiwarn monitor to ensure that it was actually happening. Now this battery-powered, handheld analyzer would be used not to test for methane and other gases in the dead ambient air of the tunnel, but to monitor the breathing air the divers needed to live.
Following Harald’s protocol, DJ used a sampling port in the plywood manifold that was jammed between the Humvee seats. Once he opened the valve of that port, it siphoned off some of the breathing air—after the air had left the mixer but just before it entered the divers’ umbilicals. DJ then held the probe from the Drager analyzer in front of the air whooshing out of that open valve. When the digital readout told him the actual oxygen percentage of their breathing air, he would call it out to Tim, who would log it on his clipboard and then phone it in to Harald.
DJ did as he was told, but the whole setup seemed bizarre to him. During his time in the Gulf, he had done plenty of deep-water dives using some form of mixed gas, whether the helium-oxygen blend heliox or the nitrogen-oxygen blend nitrox. But that gas came premixed and therefore pretested. He’d never before been asked to mix his own gas on site and breathe it, right then and there. Regardless, he’d felt burned by his previous attempts to raise questions with Harald, so he wasn’t about to do it again now.
The three divers on the excursion crew184 carried their own Drager gas analyzer. With DJ in charge of monitoring everyone’s breathing air back at the Humvee, the excursion divers used their handheld analyzer to measure the levels of oxygen and other gases in the ambient tunnel air. Even though their breathing supply was coming from their apparatus, prudence still required them to be mindful of everything in the air around them. If one of the divers encountered an emergency and felt the need to remove his facemask even for a moment, as Hoss nearly had on Monday, it would be critical to know the precise oxygen content of the ambient air. The divers had only to think back to the oxygen percentage chart they’d been given during their mine rescue training—decreased ability to perform tasks at 17 percent; dizziness and impaired judgment at 15 percent—to remember that not all oxygen-deficient environments are the same. While 9 percent would eventually lead to unconsciousness, just a few breaths at 6 percent would almost certainly mean death.
The Drager handhelds had built-in alarms that were factory-set to detect when the oxygen level dropped below 19.5 percent. Back during setup in New Hampshire185, Harald had told Tap that, because the tunnel would be so oxygen-deficient, they needed to adjust the alarm settings. If they didn’t, the device’s siren and flashing light would be going off constantly, making it hard to get work done. He had instructed Tap to lower the alarm trigger point to 9 percent. But as the divers ventured deeper into the tunnel, and the oxygen levels continued to drop, the alarm kept going off, even at the 9 percent setting. So Harald had told Tap to ratchet it down even further, to 6 percent. They might as well have set it to zero, since by keeping it at such a low percentage, they were essentially disabling the alarm.
When DJ and Tim weren’t taking their periodic measurements and calling in with their scheduled updates, the Humvee assignment could get mind-numbingly dull. Sitting in the dark, with little sound except the constant hiss of the breathing system and the muffled drips of water sliding down the tunnel walls outside the vehicle, DJ struggled to keep from nodding off. He and Tim decided the best way to stay alert was to keep talking, and they began swapping war stories. After lots of laughs, Tim grew serious at one point, asking DJ something that was gnawing at him now that he was back working in the Northeast. Just before he and Judy had moved from New York to Washington State, Tim had abandoned an old car on a Manhattan street. “You think there could be a warrant out for me because of that?” he asked DJ.
179 (Chapter 9, note 2)
about 100 pounds per square inch (PSI): Ibid., 137; Trooper Mary T. McCauley, “Interview of Harald Grob,” Massachusetts State Police Suffolk County Detective Unit, January 7, 2000, 7.
180 (Chapter 9, note 3)
gas going out into the divers’ umbilicals: Gunnar Bak Rasmussen to Tony Drybanski, June 29, 1999.
183 (Chapter 9, note 6)
There was no gauge or monitor on the mixer: McCauley, “Interview with Grob,” 5.