The welders stood at the bottom of the face and looked up. The slope above them was near vertical, roughly 2,800 feet of it rising into the Chugach Mountains north of Valdez, and the engineer Bill Howitt had already said it plainly: a caterpillar could not even work on this ground. Every section of pipe that had to go up that wall was 80 feet long and weighed about nine tons. The cable crew guiding it had nothing but a radio.
Helicopters ferried the men up to finish a critical half-mile link before winter closed in, and behind them stood a project whose owners had authorized up to ten million extra dollars a day if spending it saved a single day. Nobody needed to say what that meant. The cost of the whole line was already climbing past every estimate, and the question of whether the pipe would hold had not yet been answered. The story had begun with a road that did not exist.
North of the Yukon River, there was no highway to the Arctic at all, and nobody could haul pipe, gravel, and camps for thousands of workers across open tundra on hope. PBS called it the first job of the whole project: a supply road through the trackless northern half of the state. On the 29th of April 1974, the bulldozers went to work on roughly 360 miles of gravel road from the Yukon River to Prudhoe Bay. Alyeska, the company the oil producers had formed to build and run the pipeline, gave it one summer.
Dave Hogan, a project manager for Alyeska, had been on the job for only a week or two at a camp five miles north of the Yukon when a bulldozer worth maybe a couple hundred thousand dollars caught fire. He told Alaska Public Media years later that he could see his whole career going up in smoke. The word that came back from headquarters, as he remembered it, was that if it was in the way, they should push it over to the side and keep moving. In his words, it was going at warp speed, with crews working 24 hours a day, 7 days a week.
That pace was the whole point. Alyeska says the road took 154 days and about 3 million hours of labor, and that it was finished and dedicated on the 29th of September 1974. By the company’s count, it cost roughly $125 million. In most places, a road like that would take years to plan and build.
This one got five months. Today it is part of the Dalton Highway, which by Alaska’s figures has 20 permanent bridges and a maximum grade of 12%. The state took the road over in October of 1978 and gave the highway its name in 1981. So why the hurry?
In 1968, Atlantic Richfield struck oil at Prudhoe Bay, and PBS says geologists soon calculated that as much as 10 billion barrels lay under the frozen tundra. In February of 1969, the oil companies announced a pipeline that would cost $900 million and be finished by 1972. According to PBS, they ordered $100 million worth of steel pipe from Japan. The route had other plans.
Byron Mallott, a Tlingit leader, told PBS that Stevens Village was located directly in the proposed path of the pipeline. In March of 1970, five native villages, Stevens Village among them, sued over land they claimed along the route, and federal judge George L. Hart stopped the permit. Environmental groups went to court as well.
In Mallott’s words, suddenly the game changed. He told PBS that the native community had been facing a settlement of mere pennies an acre and no land. When Congress settled the native claims in 1971, the deal came to 44 million acres of land and close to a billion dollars, and the pipeline was still not approved. Even then, the Senate vote that shielded the pipeline from more lawsuits ended in a 49 to 49 tie in July of 1973, and Vice President Spiro Agnew broke it.
Four months later, in the middle of the Arab oil embargo, President Richard Nixon signed the pipeline into law. Then there was the ground itself. Much of the route ran over permafrost, soil that stays frozen year after year, and thawed permafrost turns to mud that cannot hold a pipe up. The cold above the ground was its own problem.
At Prospect Creek, a camp built to house pipeline workers, an official thermometer read minus 80 degrees Fahrenheit on the 23rd of January 1971, the coldest temperature ever recorded in Alaska. Every machine had to keep running, whatever it cost. Bill Fowler, an assistant project manager on the pipeline, told PBS that the project sucked up all the equipment that was available, and that he once hired aircraft and spent $10,000 to fetch a $50 part for a Caterpillar tractor just to get it started again so the work could keep moving. In other words, the flight cost two hundred times as much as the part.
The road also needed a way over the Yukon River, and the river fought back first. In August of 1974, a cofferdam failed, the temporary wall that holds the river back while crews build a pier. According to the Fairbanks Daily News-Miner, the one at the bridge had to be rebuilt. The bridge still runs about 2,300 feet.
It was dedicated on the 10th of October 1975, two months ahead of schedule by the paper’s account, and Alyeska puts its cost at about $30 million. For its first years, it carried pipeline traffic only. By Alyeska’s date, the public had to wait until 1979. And the road was only round one.
At the Tonsina River, north of Valdez, on the 27th of March 1975, the first section of the Trans Alaska pipeline went into the ground. Twelve or thirteen sideboom tractors, depending on the account, did the lifting. A sideboom is part bulldozer and part crane, with its boom swinging out over one side of the tracks. Engineering News-Record says they brought the wrapped pipe to the gravel workpad and set it in a water-filled trench up to 17 feet deep, and nine-ton concrete saddle weights pinned it down so it would not float.
According to Alaska Business Magazine, it was laid by Construction and Rigging Primary Products Corporation, a company tied to Ahtna, one of the Alaska Native Corporations created by that 1971 settlement. The native land claims that had once held the pipeline up in court had turned into companies, and one of those companies helped lay the first pipe. From there, the work spread out along the line. Alyeska says the workforce peaked at 28,072 people in October of 1975, its own staff and the contractors together.
About 70,000 people worked on the project between 1969 and 1977, and they lived in 29 construction camps strung out along the route. The camp at the Valdez Marine Terminal had 3,480 beds. The biggest camp on the line itself, at Isabel Pass, had 1,652 beds. The smallest, at Sourdough, had 112.
The rest of the numbers are hard to picture. Alyeska counts more than 100,000 pieces of pipe, 40 and 60 feet long and 48 inches across, about 550,000 tons of steel in all. At pipe yards in Fairbanks and Valdez, welders joined two 40-foot pieces into one 80-foot double joint, about 42,000 of them, before the pipe went out to the line. Crews on the line then made 66,000 field welds, moved 73 million cubic yards of gravel, and built 14 airfields.
About 2,000 contractors and subcontractors worked on the job, and the paperwork alone ran to 515 federal permits and 832 state permits. Crews also built 13 pipeline bridges, 10 road bridges north of the Yukon, and 36 south of it. Before any of that, an archaeological survey along the route cost about $2. 2 million.
The plan called for 12 pump stations, and one of them, pump station 11, was never built. Money was the reason most of them came. Al Fleming told PBS he opened a copy of Time magazine and saw a pipeline worker holding up a paycheck for about $1,500 for a single week. J.
B. Carnahan put the boom into words for PBS. It was like a gold rush. He said, “We had people from almost every country and every walk of life.
” On some nights, he said, it was like a circus, with people coming into town carrying three, four, five thousand dollars in cash. Truck driver Dave Smallwood described the roads in two short sentences. It was just bumper-to-bumper traffic. It was pretty much insanity.
Inside the camps, the money came with a routine. Welder’s helper Larry Houle told PBS it was the same wake-up time, the same bus, the same crew, and the same work site day after day, seven days a week. He admitted to PBS that he brought trunkloads of liquor into camp, and he said that in the end, he banked about $97,000 in cash. Others told PBS they did not keep much of it.
Laborer Todd Honer remembered his first landing at Prudhoe Bay as spooky, with almost nothing to see out the window because it was dark. Later, he said, came the moment he realized he maybe should have saved some of that money. Smallwood said the job did not really set him up either. He just lived well while it lasted.
Ben Holman helped move all that earth. Alaska Public Media puts the total at 93 million cubic yards, a bigger figure than Alyeska’s gravel count. Holman gave that station the plainest description of his crew: “We were dirt people,” he said. “We were building pads.
We were miners. We were mining gravel. ”
Smallwood also told PBS about the rule on the narrow haul roads. The motto then, he said, was crowd the other guy, not the ditch.
At a camp called Franklin Bluffs, he counted the people around him one day: 1,100 men, 97 women. Women were a small minority on the line. Alyeska’s fact book puts them at between 5 and 10 percent of the workforce. Diane Benson, a Tlingit woman who drove a truck on the pipeline, told PBS that camp life could mean a man walking into the shower while she was in it just to do his laundry.
She also said it was hard to keep your dignity up there, but that the whole thing was one big adventure and that she loved it. Diane Shank was 21, a graduate of Reed College with no construction experience at all, when she started as a welder helper with Pipeliners Local 798 in 1976. She told Alaska Public Media that her biggest worry was that she would not be able to do the job, and that she did not even know what a welding rod was. Years later, she testified in a lawsuit the Equal Employment Opportunity Commission brought against that union over discrimination against women and minorities.
One of that union’s welders would walk into this story at the worst spot on the whole line. Here was the problem the engineers had to solve. As PBS puts it, Prudhoe Bay oil would come out of the ground hot, and as it flowed through a buried pipe, it would thaw any ice-rich permafrost along the way. To find out where that ice was, Alyeska says engineers drilled about 3,500 boreholes and took about 15,000 soil samples before construction.
Wherever the soil was stable, the pipe was buried anyway, about 376 miles of it by Alyeska’s count. But about 420 miles crossed permafrost that warm oil would have melted. So there, the pipe went up on stilts. Howitt told PBS that every single one of the 78,000 vertical support members had a 24-inch hole drilled for it, and that a qualified person checked what came out of each hole to decide what kind of support it should be.
The drill rigs and bits arrived late, Engineering News-Record reported, and by December of 1974, 3,000 workers were drilling holes for piles and finishing concrete footings. In many of the supports, crews placed heat pipes, 124,300 in all, filled with ammonia. In winter, those heat pipes pull warmth out of the ground and keep it frozen with no pump and no power. The pipe itself rests on shoes that can slide along steel beams, and it runs in a zigzag, so it has room to move as it warms, cools, and shakes.
Nowhere did that matter more than where the line crossed the Denali fault. The United States Geological Survey says the engineers designed it for a magnitude 8 earthquake, with the ground slipping up to 20 feet sideways and 5 feet up or down anywhere inside a strip about 1,900 feet wide. At the time, it looked like one more number on a drawing. In 2002, it would decide whether the pipe survived.
So far, there was a road built in one summer, a pipe that had to stand on 78,000 legs, and a workforce the size of a small city. Now the clock started to matter. Frank Moolin, Alyeska’s project manager, had a deadline, and the Alaskan winter was working against him. In December of 1975, the cold shut down welding for six weeks.
According to PBS, 371 miles of pipe were in place by then, just short of Moolin’s target of 400. Moolin’s answer was money. Fowler put it this way for PBS: the owners told Frank that he could spend up to an additional $10 million a day as long as what he spent it on would save a day. Fowler later said that without Moolin, the pipeline would not have happened at all, and he remembered that people simply did not want to let him down.
The highest point on the whole line was Atigun Pass in the Brooks Range, about 4,700 feet up by Alyeska’s figure. There the pipe was buried in a heavily insulated concrete box. In archive footage that PBS used, Moolin describes the effort in one sentence: we have from 500 to 700 men working 24 hours a day in Atigun Pass under some extremely difficult working conditions. Then he listed them: snowfall, very low temperatures, and high winds.
Every weld on the pipeline was supposed to be X-rayed so inspectors could see flaws inside the steel. According to PBS, that meant about 108,000 welds, the double joints and the field welds together. On part of the line, that job went to a subcontractor called Ketchbaw Industries. Federal inspectors were not absent either.
In June of 1976, John W. Barnum, the deputy secretary of transportation, told a House subcommittee that his department’s engineers had made four inspection trips to the line in 1975. According to a later report by the General Accounting Office, the federal officer overseeing the pipeline had warned Alyeska again and again that year that the X-ray crews were miles behind the welders. Most people never heard what was going on between those visits.
Then Peter L. Kelly of Fairbanks, a former Ketchbaw employee, came forward. He testified before the same House subcommittee, and according to the Fairbanks Daily News-Miner, he said he had personally falsified about 30 X-rays and had watched others doing it before and behind him. That went on, he said, for about a month and a half in the summer of 1975, on more than 70 miles of section 3.
The paper quoted his reason: the reason for the falsifications was to keep Ketchbaw moving down the pipeline and keeping up with its contract. He also told the committee he believed it was an isolated case, one contractor faking X-rays to keep its contract. Fowler described to PBS what one of the subcontractors did. You find a good weld and you X-ray it 10 or 15 times from a different angle, he said, and then call it 10 or 15 different X-rays, and then say, “Well, the next 15 joints are in good shape.
Now we can move ahead and you get caught up. ” So the film in the files showed real welds. They just were not the welds the labels named. Once that came out, nobody could trust the paperwork.
PBS says that all 30,800 field welds made up to that point fell under suspicion. Investigators narrowed the suspects down to 3,955, and more than half of those were already buried. About 1,900 needed minor repair, and 37 had to be cut out and done again. By mid-October of 1976, Time magazine reported crews were down to the last 200 of those suspect welds.
In late November of 1976, the Federal Department of Transportation ruled on a waiver for the suspect welds. And in May of 1977, Alyeska was back asking for a second waiver on weld defects. Howitt told PBS what that did to the people on the job. If something as simple as an X-ray could not get done right, he asked, what else is buried?
For a pipeline that crossed hundreds of miles of frozen ground and an active fault, that was not a small question. But the line was not finished yet, and the hardest stretch was still waiting. Thompson Pass in the Chugach Mountains north of Valdez, in the fall of 1976, was the slope where this story began. In the words of the PBS documentary, the south face the pipe had to climb was near vertical, about 2,800 feet of it, and Alyeska rates the steepest grade at 145 percent.
So the crews built a tower and a winch-driven cableway, which Engineering News-Record says carried pipe sections and equipment up and spoil down. Customized tracked vehicles of 92 tons pumped fine backfill through 10-inch hoses. Elsewhere on the steep ground, bulldozers worked hanging from cables, which Alaska Public Media compared to marionette strings on giant puppets. The same report says planes bombed a mountain pass with coal dust in spring so the snow would melt faster.
Time magazine described the work on the pass as slow and hazardous, hampered by howling winds, rock slides, and blowing snow. Alyeska’s president, Bill Darch, called the push a commando raid. The pipe still had to be welded together right there on that face. One man had already been hurt by a falling rock, and most of the others decided it was too dangerous to go on.
That is when Hugh Leslie, one of the oldest welders on the job, stepped forward. He was scared, he said later, but he was going, even if it cost him his life. He told PBS he did it because it needed to be done, and that he did it for the union. The second man to go was a small welder named Richard Bane.
Bane later called it the most dangerous job he had ever seen, with rocks coming down every day and the pipe always at risk of breaking loose. For the next few weeks, in the words of the PBS film, Leslie and Bane worked in a place where only rock climbers had ventured before. Leslie worked through snow, wind, rain, and sleet. He said it felt like trying to wash the windows on the Empire State Building.
In an old recording that PBS played, his voice from back then says it plainly: no mountain like Thompson Pass. This is the granddaddy. He came down that wall, and 30 years later he was still around to tell PBS about it. The PBS film says the welders were done with Thompson Pass on the 20th of October 1976, although one PBS clip puts the last pipe there in early December.
In October, Time magazine reported that the granddaddy of all garage sales was soon to occur in Alaska. Up for sale were $18,000 bulldozers, cranes, and other pieces of heavy equipment, along with 1,500 insulated houses that had cost $10,000 each. The line was not finished yet, and the machines that built it already had a price tag. By early December, the main line was also finished through Atigun Pass.
The race was closer than it looked. The General Accounting Office later found that by the middle of April 1977, the whole project was 95. 9 percent complete against a plan of 96 percent. On the 31st of May 1977, Alyeska made the final weld near pump station 3.
Now the pipe had to carry oil. On the 20th of June 1977, at 10:26 in the morning, oil started moving south from pump station 1 at Prudhoe Bay. Two weeks later, on the 4th of July, a buried pipe and an elbow near pump station 8 cracked after supercooled nitrogen got into them. According to Alyeska, the Associated Press reported that the liquid nitrogen was about 290 degrees below zero, that it got into a pipe buried 20 feet down by mistake, and that it left several six-inch cracks.
The oil front was held at milepost 472 while crews repaired the line. The oil reached pump station 8, southeast of Fairbanks near Salcha, at 9:24 on the evening of the 7th of July. The next afternoon at 2:38, the station exploded and burned. According to the Associated Press, which reported what congressional investigators found, oil had been let through a pumping unit that was being repaired.
It rushed through an unsecured hatch into the pump building, where it caught fire. About 15 workers ran out as the oil hit the roof 45 feet up. A supervisor got on the phone to the controller and told him, “We have a fire. ” Within eight minutes, the flow through the entire pipeline was stopped.
One worker was killed, and five were slightly injured. The National Transportation Safety Board put the damage at $35 million. It blamed the failure of the poorly coordinated and inadequately supervised personnel at pump station number eight to follow precisely the written procedures for performing maintenance work and starting the pumps. It also found that nobody had sole authority over everything happening inside the station during that critical startup, that the people in the control room could not see into the pump room, and that there were no controls in the pump room to close the valves and shut the pump down.
The station was rebuilt and went back into service in March of 1978. It ran until the 30th of June 1996. By then, Alyeska says, the pipeline had delivered its 11 billionth barrel of oil and loaded more than 14,000 tankers. In 2026, a crew of 7 to 10 people started taking the old station apart, a job Alyeska expected to last about four months.
On the 19th of July, a heavy equipment accident at a check valve spilled 1,800 barrels by Alyeska’s count. And at two minutes past 11 on the night of the 28th of July 1977, the first oil reached Valdez, 38 days after it left Prudhoe Bay. The terminal there alone had cost about $1. 4 billion by Alyeska’s figure.
On the 1st of August, the tanker Arco Juneau sailed with the first load. The final bill was $8 billion, almost ten times the first estimate. According to PBS, Alyeska counts 32 fatal incidents directly related to construction, and an older Alyeska fact sheet put it at 31 people. No public list of their names could be found.
Alyeska’s motto for the job says the builders did not know it could not be done. Moolin did not get to enjoy his finished pipeline for long. He died a few years later at 48, and PBS says his friends believed his relentless style had something to do with it. Mallott told PBS that the pipeline allowed Alaska’s native peoples to become a social and economic force.
Benson, the truck driver, put the mixed feelings into words. It felt like it cut across my heart sometimes to stand on the hillside and look at what it did to this place. On the other hand, you had to appreciate what a magnificent thing it was to build this thing. We were actually proud of it.
So did the pipe hold? On the 3rd of November 2002, shortly after 1:00 in the afternoon and about 90 miles south of Fairbanks, a magnitude 7. 9 earthquake ruptured the Denali fault, right where the line crosses it. Ten days earlier, a magnitude 6.
7 quake had already hit the area. This time, Alyeska says, the rupture ran for 209 miles. The United States Geological Survey says the ground moved about 14 feet sideways and 2 and a half feet up and down, and Alyeska’s own figures range from 13 to 18 feet. Either way, the pipe slid on its shoes the way the engineers had planned.
Hillary Schaefer, an Alyeska maintenance coordinator, drove up to the fault line soon after. She found tripped anchors, cross beams, and shoes lying on the ground, but the pipe itself had settled back almost centered over where its cross beam had been. In her words, around the line, it looked like a raven had pillaged a garbage can. Mike Malvik, who supervised Alyeska’s system engineering group, said response crews were in the fault zone within an hour or two.
The line had been designed so that two supports in a row could fail, leaving about 180 feet of pipe hanging free during repairs. And in a couple of places, supports did fail. So Alyeska gathered every piece of timber cribbing it could get from the North Slope and from its partners to hold the pipe up in case of aftershocks. The moving ground had squeezed the line together like a loaded spring.
Malvik put it this way: we joke that we got an extra five feet of pipe out of the earthquake, and that the pipeline held Alaska together when the ground shifted. One of his engineers, he said, called it mother nature’s pop quiz. Malvik gave the credit to the people who drew the plans in the early 1970s. In his words, the original seismologists, geologists, and engineers did excellent work.
They expected the fault to show up in this area. Inside that zone, they used no anchors at all, so the pipe could rest on sleepers, beams set flush with the ground, and move wherever the earth pushed it. They planned out about 2,000 feet to work within, he said, and they got it right. Very few seismic engineers get to test out their assumptions and facility designs, he said, and very few get to come back and check out performance.
Schaefer added that the timing helped. Snow was late that year. The ground was not yet frozen, and crews could see and photograph the damage before winter covered it. She said she had so much respect for the design and the engineering it took to survive an earthquake that large.
Alyeska’s own damage list counts two support members bent where the pipe hit them, eight damaged supports, five damaged shoes, eleven shoes on the fault that had to be put back in place, and about 100 damaged insulation panels. Alyeska reports no leaks. The line was shut down for 66 hours and 33 minutes, and then the oil moved again. It is still moving.
In August of 2026, Alyeska says the line carried about 458,000 barrels a day. Largely alongside it, a new line is planned, an 87-mile gas pipeline 42 inches wide from the North Slope to Nikiski on Cook Inlet. The developer Glenfarne owns 75 percent of the project, and the state’s Alaska Gasline Development Corporation owns the other 25 percent. On the 30th of September 2026, the White House announced roughly $54 billion of South Korean investment in it.
Seoul’s industry minister answered that the announcement went far beyond what was agreed, and South Korea’s president said the project would only move forward if its commercial viability was confirmed and it followed Korean law. As of early October, there is still no final investment decision. One of the contractors holding a conditional award is a joint venture of Associated Pipeline Contractors from Texas, Cruz Construction, and Doyon Energy Services, a company of Doyon, another Alaska Native corporation. The men who built the first line were always in a hurry.
The ground under it never is.
