Rust: The Longest War

by Jonathan Waldman

  Then declining throughput began to affect the pig runs. The first failed run—the first which had to be rerun—was in the spring of 2001, on Neogi’s first day of work. The pig, covered in wax, hadn’t gathered enough data. To give it the best possible chance of a successful second run, controllers stockpiled the lightest, warmest crude and ran the pig in it. That rerun was the first full pig run that Neogi saw, and it was Alyeska’s last dalliance with an ultrasonic pig. Henceforth, Alyeska couldn’t get its pipeline wax-free enough.

  Resorting to an MFL pig, Alyeska suffered the same setback in March 2004. Wax ruined the run. After running a bunch of cleaning pigs, Alyeska ran the MFL pig again two months later and got enough data. Alyeska discovered thirty-five significant spots that the 1998 and 2001 pigs hadn’t picked up on at all, some of which had wall losses from a quarter to a third of the pipe. In particular, Alyeska found four dents with metal loss, one of which required immediate repair. Alyeska, though, neglected to inform PHMSA and didn’t fix it for a year. To assuage PHMSA, Alyeska reassured regulators through at least the fall of 2005 that it planned to employ an ultrasonic pig on its next run.5

  Four months later, in March 2006, external events screwed up things again. Severe corrosion in one of BP’s North Slope feeder lines caused a quarter-inch hole, and through it, five thousand barrels of crude spilled onto the tundra. Where Alaskans used to joke that BP stood for Big Provider, now they said it stood for Broken Pipeline, or Bad People, or even Bureaucratic Pandemonium. An Alyeska employee told me that BP was running a sprinkler system rather than a pipeline system. As with the Exxon Valdez, Alyeska had nothing to do with the goof-up, but it felt the ramification: heightened attention to corrosion. Under the spotlight, Alyeska decided to send a smart pig through TAPS immediately.

  That summer, Alyeska had just celebrated its thirtieth anniversary, having shipped more than 15 billion barrels of oil. That was 50 percent beyond what anyone had expected from the pipeline. Confidence outside the engineering department must have been high counting down to the pig run. Alyeska ran the pig in August. Under pressure to pig as quickly as possible, the company hadn’t prepared a cleaning regimen. In fact, Alyeska hadn’t cleaned the pipeline for most of March, following the discovery of BP’s leak. Because the BP field was shut down, throughput in the pipeline was only 450,000 barrels per day—lower than it was in March 2013. With the low flow, wax buildup must have been rapid. It ruined the run. In Valdez, techs spent a week removing wax from the pig and discovered that 20 percent of the pipe had gone unexamined.

  Alyeska ran the pig again in September. That time, in slack conditions, the pig scooted down Atigun Pass and got going so fast that the sensor heads broke when they hit welds. The pig stopped collecting data near the bottom of the pass. In March 2007 Alyeska ran the rebuilt pig for the third time—and got just enough data so that Neogi was able to combine what the pigs had gathered into a tolerable analysis.

  The next month, PHMSA fined Alyeska $260,000 for analyzing the data from the second 2004 run too slowly. The law gave Alyeska six months after an integrity assessment to determine threats to its pipeline, and the company had taken eleven. Alyeska objected, citing “technical difficulties” and electronics failures caused by wax. PHMSA lowered the fine to $173,000. Still, it was the largest civil penalty levied against Alyeska in a decade.

  The architect of the failed pig runs in 2001, 2004, and 2006–07 was a large churchgoing engineer named Dave Hackney. He’d come to Alaska from New York in 1962, studied mining engineering in Fairbanks, and helped build the pipeline. In 1983, from Alyeska’s quality assurance department, he volunteered to oversee smart pigging when nobody else did—thinking it was good job security. For years he wore a hard hat that said “Swineherd.” Like many fastidious engineers, he maintained a Teddy Roosevelt mustache and kept a stable of pens in his shirt pocket. He claims to have taught Neogi much of what he knows. At Alyeska now, his name is verboten, and his last runs are infamous. “Previous ILI engineer” was how people referred to Hackney. In 2008, after Alyeska brought in a new president, Hackney was transferred out of the pigging group and soon thereafter left the company. He was sixty-two. “The official line is ‘I retired,’ ” he told me. He sued his former employer and settled the dispute out of court. Now neither party can criticize the other or get into the matter. Hackney laid blame on tough circumstances and “the politicians that won’t let us drill for more oil,” and insisted that on TAPS, pig reruns aren’t failures (millions of dollars notwithstanding) so much as par for the course—and that the result is still good corrosion data. He left it at “Whatever happened, happened.” He remains proud of his work defending the pipeline.

  Neogi had some sympathy for Hackney, without going so far as to call him a victim. “Who knew there’d be this much wax?” he said. “It’s not like we ever half-assed it.” At the same time, he didn’t think events concluded in such a crazy manner. “It’s a big deal. You wanna make sure you’re taking care of the integrity of the pipeline. The definition of insanity is doing the same thing again and again, and expecting different results. If I do this three times and I fail, I hope Alyeska sees fit to bring in someone who has better ideas. I don’t think it’s personal. It’s an important task, and Alyeska wants to make sure it gets done right. Dave, Elden, myself—we’re replaceable. The pigging is not.

  “In 2006, we learned a lot,” Neogi added. Alyeska learned to run a much tighter ship, and Neogi learned not to hurry the process. It’s the same thing he learned with his fish.

  MILE 800

  Twenty-four hours before the tool was due in Valdez, Neogi left Anchorage in his wife’s minivan, the one with the B TEAM plates. He drove to Valdez not just because he hates flying but also because a storm across most of south central Alaska had grounded flights into Valdez. He had to get there, so he was driving three hundred miles through whatever awaited. Conditions on the first hundred miles of the Glenn Highway were bad. The Matanuska Glacier and the dramatic peaks of the Chugach, off to the right, were obscured in white. It took three hours to get to Eureka Summit, where conditions improved. Here Neogi—who’d been snacking on dried fruit—stopped for lunch. He was wearing a red T-shirt, blue jacket, green track pants, and sneakers. He was sporting a gold chain, the big watch, and two rings. He looked like he was on vacation, and not in Alaska. From two snow machiners at a nearby table, he drew a couple of looks.

  Neogi sat down, ignored the bottomless twenty-five-cent coffee, and ordered a Eureka burger, pronouncing it “you-wreck-uh.” He checked his BlackBerry, and had received twenty-five emails. One informed him that the final cleaning pig had recently arrived in Valdez. It was there, but it was still in the receiver. He called an engineer on his team and said, “Call me when the sweep tool comes out, and let me know how much wax there is.” After eating quickly, Neogi bought a pack of gum. For him, it was the equivalent of a pack of cigarettes.

  Sixty miles down the road, in Glenallen, Neogi took a right on the Richardson Highway, and paralleled the pipeline as he headed south. As he drove, he evaluated the pig run and considered the various hiccups encountered: the stuck valves, the helicopters, the tracking, the flow variation. He said that the most significant was the current twenty-one-hour separation between the sweep pig and the smart pig. He didn’t like the possibility between the two for wax buildup. He held the index finger and thumb of his left hand a hair-width apart and said, “If the wax even builds up only one millimeter . . .”

  He restarted. “I would like to have had more oil in the pipe. Quantifying wax is very difficult. There’s no pure science that says this is it.”

  Then he said something surprising: “Even in engineering, there’s things that are half magic, half repeatable.”

  Beyond mile 700, it started to snow again. Neogi insisted that he was relaxed. He insisted that even if the smart pig did hit a check valve, the pig’s cups, rather than its electronics, would take the blow. He insisted that the pig wouldn’t find any corrosion below the
clamps on the stilts. “You just don’t know,” he said. “You do the best you can with what you know. I don’t worry about it, because I’ve done my homework. There might be a little hiccup here or there. My game plan has always been to study the hell out of the thing. Where can you get better? If something happens that you haven’t thought of, then you just put your hands up and say, ‘Well, this one got by me.’ ” Ten minutes later, after a few glimpses of the Wrangell Mountains poking out from clouds to the east, Neogi asked me if a cold front had come in. Cold wasn’t necessarily bad news, because frozen soil conducts less heat from buried pipe than does wet soil. Either way, at the start of the pig run, Neogi was thinking about delta P. Now, approaching the end, he was thinking about delta T.

  Eighty miles from Valdez, as he was descending to the Little Tonsina River, Wasson called. It was 2:47 p.m. Neogi pulled over and put on flashers to talk to him. Neogi’s half of the conversation went like this:

  “Hey, Ben. You got five barrels of wax? . . . Yeah, but that seems high . . . What did the one before bring in? . . . Is it hard wax or soft wax? . . . Hmm . . . Well . . . You don’t have any pictures? . . . Okay . . . Okay . . . Okay.”

  After Neogi hung up, he looked in the rearview mirror, pulled a little farther off the shoulder, and called his boss to report the news. When he hung up, he said, “Well, there’s a surprise.” A couple of barrels, he admitted, was what he was hoping for. Then he qualified it. “Five’s not bad—it’s not like forty,” he said. Shortly thereafter, he approached Thompson Pass, where oil in the Trans-Alaska Pipeline cascades almost a half mile down the Chugach en route to the sea. Down the stretch, recent pigs have not been able to capture any data, and Neogi hoped the pig wouldn’t exceed fifty miles per hour. As he neared the pass, he started humming Beethoven. It was the only other time I heard him do it.

  After checking in to his hotel room in Valdez, Neogi spent the evening watching the NCAA basketball championship game. Then he had dinner at the only Thai restaurant in town. It was also the only place with an aquarium full of fish. Over dinner, he insisted that he’d sleep okay, that he had learned to deal with butterflies in his stomach long ago in badminton competitions, and that the last time he experienced the feeling was when his kids were born. He said that he knew that he had the most to lose but that he was also well prepared. He said it was like being in sight of the finish line at the end of a marathon. Crossing the line is all he wanted to do. He denied that crossing it would be exciting. “You can’t be in this business if you’re looking for excitement,” he said. “Think of a war general: he can’t be running around, going nuts at every little thing. He has to be calm and collected.” In that manner, Neogi said he’d return to his room, call his wife and kids, read something, and—twelve hours before the pig arrived—sleep just fine.

  Overnight, a foot of snow fell, and in the morning it was still snowing. The Valdez Marine Terminal may be the northernmost ice-free port on the Pacific, but it occupies one thousand of the most precarious and snowiest acres anywhere. Snow rose to the height of stop signs, and higher. Berms and snowbanks were so steep and gnarled in places that they resembled seracs; the streets veritable crevasses. One winter brought forty-six feet. The immense depth was more astonishing than the immense width of the North Slope. Sliding off of the roof, snow somehow broke a window on the fourth floor of Alyeska’s modern glass office building. The famous statue honoring the (mostly) men who built the pipeline, with the plaque that reads “We didn’t know it couldn’t be done,” was so covered in snow that it couldn’t be seen and couldn’t be gotten to. All over the terminal that morning, loaders and plows were crawling and beeping. In front of the building, men were shoveling.

  Inside the building, Devin Gibbs sat in the cafeteria, ensconced in his huge blue parka, eating breakfast alone. It was his thirty-sixth birthday. The pig was due in an hour. He said, “The best birthday present would be if the pig comes in clean.” Two flights up, Scott Hicks, the trim, affable director of the terminal, was walking the halls. BlackBerry in hand, gray V-neck sweater over a blue-collared shirt, he said, “I just hope it comes in in one piece.” Behind wire-rim glasses, he winked. Ten minutes later, Dave Benes, the lead technician who was about to remove the pig from the pipe, insisted that the action would be standard operating procedure. A big guy in denim coveralls, he had a confident mustache and goatee beneath a camouflage Alyeska cap. “Yard ’er out, get ready for the next one,” he said. Neogi, who appeared preoccupied, said very little.

  The pig receiver was just up the hill, in an innocuous tan warehouse called the East Metering Building. On the north wall of the building, there was a tall bay door and a rusty human-sized door beside it. A sign reading DANGER: MEN WORKING OVERHEAD had been duct-taped above two others that said KEEP THIS DOOR CLOSED and HARD HATS REQUIRED. On the east wall of the building, about ten feet up, just poking out above a snow bank, there was a little orange milepost sign. It said 800. The end of the pipeline was just inside.

  At 9:05 a.m., the bay door rolled up. Inside, some familiar features were visible: a big yellow overhead crane, the little red gantry crane. Below them was the end of the pipe. Aside from THE END stenciled onto it, the end of the pipe didn’t quite look like the end of a pipe—because four smaller branches protruded from the stub. They took the oil to the terminal’s eighteen enormous tanks, from which it could be loaded into thousand-foot tankers at the berths down the hill. But that stub was it: the end of the line. In front of the door on the end of the stub, taped to the concrete floor, were two long layers of white oil-absorbing cloth. Beside the white carpet were black drums, lined with plastic bags—for shoveling wax into. There were eight of them. Beneath the floor, lodged somewhere in a section of concrete-encased pipe, was half of a pig that had been ingested in 2012. It had since been named Theodore.

  At 9:10 a.m., Neogi, Wasson, and Gibbs walked in from the east door. Ten minutes later, Benes and three other station techs, in matching brown coveralls, hard hats, steel-toed boots, and gloves, entered through the bay door. One had a wrench in his pocket. Another had a radio strapped to his chest. The pig was less than a mile away.

  It was April 9, 2013—twenty-six days since the pig was launched from Pump 1. During this span, the length of a day in Valdez had grown by nearly two and a half hours.

  At 9:47 a.m., the pig arrived.

  Last time anyone saw it, the pig was sixteen feet long. Nobody knew if it had been mangled or compressed. The pig trap was just under forty feet long, but the final six-foot section was a deadleg. To receive the pig, the techs planned to stop it in the thirty-four feet of the trap through which oil flowed. With the pig parked there, they could use the oil flowing past it to flush wax off of it. Then they would divert the oil, close the valves, drain the trap, open it, and yank out the pig. Between the trap door and the massive ball valve at the upstream end, they didn’t have much room.

  To get the pig to come in slowly (they didn’t want to slam the $2 million device into the trap door), the techs slowly opened another gate valve at the far end of the building. As the valve began to open, oil behind the pig found an alternate route to the four tank lines, leaving less oil to propel the pig forward. Tracking the pig through the final hundred feet of pipe was easy. Screws placed on the pipe fell as the pig passed. Geophones caught the sound of the pig’s movement. Receivers picked up the signals from the transmitters in the pig. Then the pig passed the ball valve, entered the receiver—which was really a pipe within a pipe—and tracking the pig became much more difficult. The screws no longer responded, because there were now two inches of steel between them and the magnets on the pig. The signals from the transmitters vanished. The geophones revealed nothing. The techs opened the gate valve a little bit more, until it was a quarter open, and enough oil took the new path that the pig came to rest. Where was anybody’s guess.

  Gibbs checked for transmitter signals, but they were null. He felt blind. It was the most difficult pig trap he’d ever seen, and he’d see
n thousands. His only useful tool was an old-fashioned gaussometer. Registering the magnetic fields around the pig’s magnets, it revealed the polarity switch between the two banks, giving Gibbs a six-foot window of precision. Wasson, ever the surveyor, used a plastic Boy Scout compass to the same effect. He and Gibbs reached the same conclusion: the pig had come to rest closer to the ball valve than to the deadleg. Gibbs was pretty sure the pig was clear of the valve, but to be absolutely sure, he needed three more feet. “You don’t want your tail hangin’ in the valve,” he explained later. He’d never shut a valve on a pig, and wasn’t about to by making any rash decisions.

  Neogi agreed with the logic and came up with a plan to push the pig a bit farther forward. He had operators stack up oil in Thompson Pass and then let it go.

  When they next measured, a dozen men participated. Their conclusion was unanimous: the pig had moved a dozen feet forward and was squarely in the receiver. Using a red paint marker that he had brought with him, Gibbs drew a little vertical hash mark on the pipe at the spot where the pig’s magnets were centered. A measuring tape confirmed that from there, the pig was well clear of the ball valve. At 11:10, the techs opened the gate valve all the way. Neogi had expected to see the pig by now, but he had to wait a bit longer. It was time for lunch.

  At noon, techs closed the valves on either end of the receiver, isolating the pig inside. They closed the bay door, activated the ventilation system, and began draining the trap. All told, at least fourteen men were in the building, more than twice the normal number. The Valdez crew was responsible for draining the trap, opening the door, removing the wax, and closing the door. The Baker Hughes and crane crews were responsible for lifting and moving the pig. Neogi and his integrity management crew were pretty much in the way, like a bunch of expectant fathers. But it was their project, so there they were.