Rust: The Longest War

by Jonathan Waldman

  “I want to know exactly what’s going on,” Neogi said. He had dark eyes, short, dark hair with flecks of gray on the tips, and was wearing a navy blue fire-retardant shirt. Built like a soccer player, he spoke in a quick staccato, with a voice that was reedy and delicate, and with a noticeable Indian accent. “The great fear is that there’s some corrosion that we can’t see,” he said. “If you can’t see it, you can’t do anything about it because you don’t know it’s there. Whatever we can assess with our pigging, we can fix.” His manner was formal but blunt, almost brusque. From the pig, he said, he yearned to get what he called the pulse of the pipeline. “You break your tool, and you don’t get data,” he said. “We’ve done so much preparation—everything we can—but it’s possible we’d have to redo the run.” This wasn’t idle conjecture. On account of wax and low flow rates, in the last dozen years, half the smart pig runs have failed.

  That afternoon, shortly before Neogi began humming, the technicians spent forty-five minutes testing the pig’s lithium batteries, each of which weighed twenty-five pounds, cost $3,000, and was not rechargeable. They checked the hard drive in the rear of the pig, and they repositioned one of the pig’s two transmitters, which would make tracking it during its eighteen-day journey possible. Meanwhile, Neogi and several executives from the pipeline’s operator, the Alyeska Pipeline Service Company, ran through the second of three final “go/no-go” meetings. Though pigging the pipe is a routine maintenance procedure, it is by no means casual: hazards abound in opening up the end of a vessel full of pressurized crude, putting a solid object in a pipeline designed for the transport of liquid, and trying to follow said object through the Arctic in the winter. Plenty could go wrong. In one of the more recent pigging mishaps, known as the gas excursion incident, Pump Station 1 nearly blew up. That was BP’s fault, but Alyeska took heed. More recently, a pig was sucked into a relief line at a pump station midway down the line. That the relief line was only sixteen inches in diameter, and guarded with pig bars, was not a sufficient deterrent to the forty-eight-inch pig. This has happened at least a half dozen times. When it happened in 1986, and the pipeline was shut down while the pig was extracted, that meant more than a quarter of the nation’s oil wasn’t moving toward California. Pigs have made it all the way to Valdez, Alaska, only to be ingested in relief lines there. Other pigs have damaged the pipeline, or gotten stuck in it and been destroyed during their extraction.

  Discussing conditions, logistical readiness, and safety, a handful of directors and vice presidents now considered whether or not to run this pig. Over the phone, Neogi reassured them that everything was on track. Hydraulic procedures for getting the pig smoothly down the slack mountain passes were ready. All other pipeline maintenance had been halted. Up and down the line, all eyes were on the pig. It still had a green light.

  That evening, after one final check, the Baker Hughes technicians turned on the pig and then, using a pair of overhead cranes, loaded it into the launch tray at the end of the pipeline. A half dozen pump station technicians assisted, as did Neogi. They wore Tyvek protective suits and gas masks and worked slowly and carefully. They called the pig “the tool.” When they opened the end of the pipeline, vapors seeped out. Gas detectors blinked, and explosion alarms went off. The ventilation system whirred. A stubborn beast, the five-ton tool did not make things easy. Just sitting in the launch tray, it required hydraulic jacks to support it. Shortly before midnight, the men finished loading it, and then closed and secured the end of the pipeline shut. Then they waited. They planned to launch the tool at seven in the morning, exactly twelve hours behind a red urethane pig of lesser intelligence. That pig, like a giant squeegee, was scraping the line clean. It was the last of nine such scraper pigs that, by Neogi’s design, had been shoved down the pipeline in the previous six weeks. Neogi had kept track of how much wax these pigs had pushed out in Valdez, and graphed it. From 1,200 pounds, the mass had dropped to 400. The line was as clean as it was going to get, primed for inspection. It was ready for the smart pig.

  Neogi woke at five o’clock. On the drive from Deadhorse to Pump Station 1, the flares of distant oil fields were barely visible in the dark. He drove through two security gates. Next, he plugged in his truck beneath the ice-encrusted radio tower, walked through the refrigerator door, and took off his fur-lined parka. Then he called Alyeska’s main office, in Anchorage, to participate in the final go/no-go meeting. The men talked slowly. A storm seemed to have passed; the windchill was no longer 50 below. Their paperwork was in order. No new safety concerns had emerged. For twenty minutes they deliberated, and Neogi was never surprised—a good sign. The pig launch was on. Nobody noticed that it was the Ides of March.

  A few minutes before seven o’clock, three station technicians opened valves in the launcher and diverted oil behind the pig. It didn’t budge. The techs, all mustachioed, sent more oil behind it. Still it wouldn’t budge. The pig—the biggest and most capable that Baker Hughes makes—was by far the heaviest that had ever been put in the Alaska pipeline, and the entirety of the pipeline’s flow pushing against it didn’t compel it to move. In the control room, where Neogi watched a little red light on a panel that would indicate the pig’s departure, he slouched, biting a fingernail on his left hand. His shirt was untucked. He looked like a family member in a hospital waiting room, edgy with anticipation. That this was the first pig under his management also contributed to his anxiety. Over the radio, he listened as the techs requisitioned more oil from two giant tanks. Up the flow went, to 600,000, 700,000, 800,000 barrels per day, and still the pig didn’t budge. It didn’t budge until propelled by 865 pounds per square inch of crude, flowing at a rate of 840,000 barrels per day—more than half again the typical flow of oil in the pipeline. It finally took off at a quarter after seven, bound to see what rust had done to the biggest, baddest oil pipeline in the world. As the pig headed south into the barren, wintry immensity of Alaska’s North Slope, it sounded like a train. All Neogi said was “All right.”

  For two decades, the Prudhoe Bay oil fields—Sadlerochit, Northstar, Kuparuk, Endicott, Lisburne—have been declining steadily. Yearly, immutably, they produce 5 percent less oil. The result is that TAPS now carries one quarter of the oil it was designed to carry. It comes out of the ground colder than ever and flows more slowly toward Valdez. Crude used to make it to Valdez in four days, as if running seven-minute miles. Now it walks. En route, it cools off even more and, as it does so, deposits more wax on the pipeline. A doctor would call the pipeline arteriosclerotic. While a pipeline waxes, its diameter wanes. Declining throughput makes things difficult for Neogi, but it makes them even more difficult for agencies estimating the pipeline’s lifespan. The pipeline was designed to survive as long as the oil fields. Lest it clog, it must stay warm, which means that it must remain full of flowing oil. In a perverse symbiosis, the pipeline needs the oil as much as the oil needs the pipeline. As a result, while the consortium of agencies that oversees the pipeline has written that it “can be sustained for an unlimited duration,” Alyeska figures that it’ll survive until 2043, and the state of Alaska figures that it’ll expire a bit sooner. Private consultants, hired to estimate the life of TAPS, mention only “the future” and write of “diligent upkeep” in passive sentences. The estimates all couch what nobody wants to say: the pipeline, once the largest privately funded project in America, and one of its greatest engineering achievements, is now an elderly patient in intensive care.

  The companies that built the pipeline foresaw such a future and tried to avoid it. In the immediate aftermath of their 1968 oil discovery, they considered every alternative to a pipeline. They considered extending the Alaska Railroad to the North Slope, until they realized that it’d take sixty-three trains, each one hundred cars long, every day, to ship their oil. They considered trucks, calculating that they’d need nearly the entire American fleet in addition to an eight-lane highway. They looked into jumbo jets supplied by Boeing and Lockheed, turning away when it became
apparent that their air traffic would exceed the combined air traffic of all the freight in the rest of the country by more than an order of magnitude. They looked into blimps. They commissioned the world’s largest icebreaking cargo ship, and after it got stuck in the Northwest Passage, they seriously considered using a fleet of nuclear submarines to ship the oil, under Arctic ice, to a port in Greenland. Reluctantly, out of alternatives, they settled on a pipeline. A steel tube winding across Alaska was ten times the rust risk of a giant copper lady standing in New York Harbor, and they knew it.

  On most other pipelines, “events” or “incidents” or “product releases”—what the rest of us call leaks or spills—are most often caused by third-party damage. By this, the industry means accidents. Heavy equipment is usually to blame; pipeline ruptures are most often caused by collisions with bulldozers and backhoes.1 On TAPS, since there’s so little construction across the vastness of Alaska, the risk of accidental third-party damage is low. Natural hazards, on the other hand, present threats in abundance. Earthquakes, avalanches, floods, and ice floes all threaten TAPS. But what really keeps Alyeskans up is corrosion. It’s the number one threat to the integrity of the Trans-Alaska Pipeline, enough to make engineers in the last frontier dream of Bakersfield.

  On account of that threat, the pipeline was outfitted with the greatest corrosion-protection features of the era. Its principal protection was its coating: paint. As a backup, a zinc strap the size of a wrist (a giant anode) was buried under the pipe. Though TAPS was, boldly, called rustproof, the defense proved insufficient. Like all coatings, the one on TAPS proved vulnerable—but Alyeska didn’t learn quite how vulnerable for a dozen years. When it did, the company beefed up the pipeline’s corrosion protection with ten thousand twenty-five-pound bags of buried magnesium anodes (recall that magnesium is the least noble metal, most willing to sacrifice itself), and a cathodic protection system consisting of a hundred-odd rectifiers spitting a low voltage into the pipe. Unlike the buried zinc strap (the condition of which and its very existence are mysterious), the mag bags and the cathodic protection system are testable: corrosion engineers can disconnect them and use probes that measure the changes in the voltage in the soil. Because rocks resist current, the cathodic protection system doesn’t work well in rocky areas, leaving corrosion engineers to their final tool: coupons. On the pipeline, a coupon is a one-inch square of steel, connected to it and buried along it, serving as a surrogate. Alyeska has about eight hundred of them. But coupons don’t prevent corrosion; they just help engineers monitor it. In a way, monitoring is Alyeska’s second line of defense, and Alyeska does a lot of it.

  Like all major pipelines, TAPS is monitored by leak-detection software, which compares the flow of oil going into the pipeline with the flow coming out the other end, and also scans for sudden pressure drops. But unlike other pipelines, it is also monitored regularly by pilots using infrared cameras to hunt for signals that the hot oil has escaped into the cold Alaskan earth, as well as by “line walkers” who hunt for dark puddles and squishy tundra along the pipeline, and by controllers watching an array of hydrocarbon-detecting and liquid-detecting and noise-detecting sensors shoved into the ground alongside it. And then there are the dozen state and federal agencies looking over the shoulders of the thousand people operating the pipeline, making it the most regulated pipeline in the world. But because a smart pig is the only way for Alyeska to determine if its pipeline is about to spring a leak before it has actually done so, and because Alyeska operates under more regulatory scrutiny than any other operator, it sends smart pigs down the line nearly twice as often as any other pipeline operator. It employs a smart pig once every three years, and has been doing so since long before federal pipeline laws stipulated it.

  Thanks largely to smart pigs, TAPS hasn’t suffered a corrosion-induced leak since it began operating in 1977.2 Over its first thirty years, Alyeska reviewed nearly 350 potential threats to the pipeline, including dents, wrinkle bends, weld misalignments, ovalities, gouges, and corrosion pits. The majority of these problems were found with smart pigs. But pigging hasn’t been a cakewalk for Alyeska. Early smart pigs weren’t so bright or amenable, and since 1998, smart pigs have been stymied by wax. In fact, wax precluded Alyeska’s preferred smart pigging technology: ultrasonics. Such pigs gather data by bouncing sound off the walls of the pipe and listening for echoes. The ultrasonic method, a direct measurement, is superior to magnetic flux leakage (MFL), which is indirect and inferred. After 2001, ultrasonic pigs couldn’t gather enough data. The wax blocked the sound. The effort was as futile as trying to get a fetal ultrasound from a pregnant woman wearing a sweater.

  Consequently, keeping the pipe clean has become a priority nearly as great as keeping it whole, because the latter depends on the former. To keep it clean, Alyeska sends cleaning pigs south weekly. The company keeps a fleet of a dozen such pigs at a maintenance yard in Valdez, and in a perpetual relay, these pigs go back and forth: up the haul road, down the line. The managerial pigs—the smart ones—wait patiently while these janitorial pigs stay busy. Before the last smart pig run, Alyeska sent a janitorial pig south every four days for a month. When these pigs pop out in Valdez, they usually push out ten or twenty barrels of wax. In the pig mobile, they go straight to the pig wash. The wax, a hazardous material, is collected in barrels and shipped out of state. Once, not many years ago, after the pipeline wasn’t pigged for six weeks, a pig pushed out forty-seven barrels of wax.

  Beneath all that wax, on account of corrosion, the one-billion-pound pipeline loses in the vicinity of ten pounds of steel a year: the same as an old Ford. Most of that metal loss is on the outside of the pipe, where it’s buried. The inside is, well, nicely oiled. The exception is inside pump stations, where the pipe branches through valves and turbines. In deadlegs—hydraulic culs-de-sac, where oil sits stagnant—microbial-influenced corrosion is a threat.

  If corrosion struck uniformly, such that the pipeline lost metal evenly and consistently, maintaining it would be vastly easier. After a thousand years, 99.999 percent of the pipe would still be there, sans weak spots. But rust doesn’t work like that. It concentrates in relatively few places, begetting more rust. Alyeska responds only to those places that present severe integrity threats. It looks at spots where 35 percent or more of the pipe’s wall thickness is gone, and where metal loss leaves the pipe at risk of bursting, which it determines from a formula developed by the American Society of Mechanical Engineers.3 In pipeline lingo, this response is known as intervention criteria. Alyeska’s are stricter than most. The company also gives itself a safety margin. It assumes that tiny pits are six inches long, and it assumes that the pipeline is operating at maximum pressure, when, in fact, as it runs out of oil, it is not. Depending upon where those spots are located—whether they’re in sensitive or populated areas—Alyeska assigns each significant threat a PYTD: a Potential Years to Dig. The severest corrosion threats earn a 0 and get dealt with immediately. The mildest get 8s, or 15s, or even 29s. In this manner, even though Alyeska may address a dozen repairs a year, a backlog accrues. Examining the backlog, it is easy to be alarmed by not-infrequent reports of pits half the thickness of the pipeline.

  Neogi exhibited no such alarm. “Take an average pipeline—Williams, Enbridge, BP—we’re in pretty good shape,” he said on the day he launched the smart pig. “TAPS was really well thought out. They designed a thing that’ll stand the test of time.” About maintaining the Alaska pipeline, Neogi made two analogies. “People think that a vehicle can last only ten years,” he said, “but that’s not true. If you have ten people working on it—” He dropped it for a better one, comparing the pipeline to a painting at a museum, where curators, monitoring humidity and light, keep masterpieces pristine for hundreds of years. “The only thing that would stop us,” said Neogi, “is the flow of oil . . .”

  MILE 104

  In every direction, all that was visible was white and blue. The North Slope was as flat as Kansas and just as tr
eeless, but a good bit colder. It was cold enough to freeze pen ink. The only break in the flatness was a couple of stalagmitic ice mounds, called pingos, sticking up like erratics. The sun, low on the horizon, reflected off the frozen Sagavanirktok River and made mirages out of the distant Franklin Bluffs. The pipeline, zigzagging along, appeared inconsequentially small. Beside it was an orange snowcat, a treaded vehicle common at ski resorts, and inside, behind frosty windows, were two men, waiting for the pig.

  Years ago, when pigs hustled along more rapidly, it was possible to hear them pass above ground or below. One employee, at Alyeska since the beginning, said pigs sounded like jets in the movie Top Gun: sheeeeooow! A man who didn’t trust his ears as much could balance screws on the pipeline—at least where it was above ground—and wait for the magnetic field of the pig to knock them over. Or he could jam a screwdriver into a part of a valve along the pipeline and listen to vibrations in the handle. Such behavior didn’t last long in the Arctic. Now these men listened for the smart pig with a geophone: essentially, a big, amplified stethoscope.

  To do so, they strapped a metal probe onto the biggest appurtenance of the pipeline: in this case, the split ring of a flange. To block the sound of the wind, they wrapped the probe in a sheet of oil-absorbing cloth—what they called a sorbie—and duct-taped it in place. Wired to the probe was a battery-powered radio transmitter, which they set in the snow. Then, rather than stand out in the cold, they sat in their snowcat, turned the radio to 107.1 FM, and listened to the vibrations in the pipe while keeping their ears warm. They turned up the bass and turned down the treble. Since the FM dial was empty across the vast range of the North Slope, 107.1 broadcasted local static. When the men threw snowballs near the pipeline, the radio chirped. When they tromped around in the snow, it was as if Indians were on the warpath. Thunk thump thum thum thump. They set up a half hour before the pig was due and then sat quietly, with the engine of their small machine turned off. In the giant Arctic, they were listening to a radio station that didn’t exist, looking for something that you couldn’t see, which in turn was hunting for something that didn’t belong. As the pig approached, it produced a rhythmic beat as it regularly hit welds along the pipe. When the pig went by, it whooshed unmistakably, unlike any standard radio song.