Who Killed Bertha? Why the Seattle Tunnel Disaster Isn’t What It Seems, and Why WSDOT Is Making a Dangerous Bet

Two workers, who request anonymity, suggest the tunnel-boring machine was pushed beyond its limit by a foolhardy contractor.

April 18, 2018By Leslie Helm


Berthas in heavy rotation (above): A giant red crane slowly rotates the SR 99 tunneling machines front end from a horizontal to a vertical position as it is lowered into the access pit on Aug. 24, 2015, two years after the incident occurred.

This story follows our feature story, “Spin Zone,” which appeared in the October 2017 issue of Seattle Business magazine. Click here for a free subscription.

Sometime during the midnight shift on Wednesday, December 3, 2013, just about 1,000 feet into a planned 9,270-foot journey, Bertha, the tunnel-boring machine (TBM) made by Japans Hitachi Zosen, smashed through a 120-foot-long steel well casing, pushing it high above the ground.

Bertha, the largest TBM ever built, raising the bar, wrote Juan Louis Magro, a construction manager for Seattle Tunnel Partners (STP), the joint venture of Spain-based Dragados and California-based Tutor Perini contracted to build the Alaskan Way Viaduct replacement tunnel. Isnt that power or what? Magro wrote, punctuating his email with a smiley-face.

The high spirits were short-lived. Three days later, Bertha ground to a halt. Investigation later revealed that its seals and bearings had been damaged from overheating. Repairs led to a two-year delay in the $3.2 billion project.

What happened during those three days is at the center of two separate court cases that will determine who will be responsible for the $500 million in additional costs incurred as a result of the badly damaged machine.

Two workers involved in operating the machine for STP during that period recently contacted Seattle Business magazine and offered details that shed new light on what happened, possibly helping to frame the arguments that will be made in the two court cases.

The workers had read Spin Zone, which Seattle Business magazine published in October. For that story, we spoke with executives and workers at Hitachi Zosen in Seattle and Japan to provide a perspective that challenged the prevailing view that Hitachi had somehow failed by building, in the words of Washington Governor Jay Inslee, a Datsun when something far more powerful and reliable was required. Hitachi argued that the problems were caused by a combination of things, a kind of perfect storm that included Berthas running into the steel well casing as well as operating errors by STP.

Since neither the Washington State Department of Transportation nor STP agreed to speak to Seattle Business, our story depended heavily on Hitachis perspective. The two workers who read the story, however, offered additional details that confirm and clarify many of the issues addressed, issues that will be at the center of the two upcoming court cases.

The first case, which will come to trial in King County in October, is between STP, which sued its insurers for $80 million it believed it was due on an insurance policy for which it paid a $25 million premium, and the insurers, who countersued STP and Hitachi, claiming the two had conspired to blame the steel well casing for the damage to Bertha even though they knew that wasnt true. (STP also sued Hitachi in the hope of covering cost overruns in the event the insurers wouldnt pay the claims.)

In an odd turn of events, WSDOT is making a dangerous bet by siding with the insurers against STP and Hitachi, thereby potentially forgoing the $80 million that would otherwise help pay for the delays. WDOT apparently fears that a successful claim by STP against its insurers would strengthen STPs position in a second case to be argued in Thurston County, likely sometime next year, which will determine who should cover additional costs of a half-billion dollars related in large part to Berthas breakdown and resulting delays.

Some experts, including one hired by Berthas insurers, appear to support Inslees claim that Bertha wasnt powerful enough until Hitachi rebuilt some portions of it during the shutdown. It seemed inconceivable to them that a steel casing 8 inches in diameter could stop a machine weighing 7,000 tons, measuring 58 feet in diameter and pushing with 31 million pounds of thrust. Some argued that Bertha, at the time the largest boring machine ever made, had been badly designed and was too unwieldy to control.

The two workers, who asked not to be named for fear of retribution, ridiculed Governor Inslees contention that Bertha was like an underpowered Datsun. [Bertha] was no Datsun, said a worker well call Jim, who had worked as a boring machine operator at multiple tunneling companies for several decades. Bertha would move through the ground fast. It had power out the yin-yang.

As for the argument among some experts that Bertha was too large and hard to control, Jim declared, That machine could cut and steer like a dream. It was very responsive. The second worker well call him Paul also had operated Bertha and agreed.

Heres what happened based on their accounts:

Operating the machine was a stressful job that required full concentration. There were three men in the control room operating a machine roughly the size of a basketball court. One operator was at the main controls, a second operated devices that sprayed grout to fill open space above the curved, interlocking concrete segments that Bertha placed in a ring to extend the tunnel six feet at a time. That worker also added the conditioner that had to be added to the soil to create the toothpaste-like consistency required for it to be pulled out of the ground and transported by conveyor belt to trucks waiting at the back of the 326-foot machine.

A third person rode shotgun, helping the operator keep an eye on the machines 10 or more gauges. Without any view of where the machine was headed, the operators regarded these gauges as their eyes and ears. They included a laser beam that kept the TBM on track, gauges to ensure pressure on the cutter head was balanced by pressure in the chamber behind the cutter head (critical for an Earth Pressure Balance machine like Bertha) and a video feed of the conveyor belt showing the operator what kind of dirt was being pulled up from the chamber by the ribbon screw so that operators would know what kind of soil conditioner should be added.

The operator drove Bertha forward by manipulating 56 hydraulic rams braced against the concrete tunnel rings behind the machine and pushed the front of the boring machine into the face of the tunnel so its 23 750-horsepower motors could turn those cutters with just the right amount of pressure to scrape away at the face of the tunnel. Yet another gauge showed how fast Bertha was advancing in millimeters per minute.

In every tunnel dig, the workers said, the first 600 feet or so they were adapting to their new TBM, learning how to operate it and making adjustments. Since Bertha was the largest TBM ever built, says Jim, its not surprising that we had to deal with some bugs, some electrical issues that had to be worked out. Another issue was that everybody working on Bertha developed a sniffly nose and a lot of people were getting sick, said Paul. He figures that was from going through landfill, which hadnt been touched since 1887 and had trapped all manner of gases.

By December 2013, Bertha had left those issues behind it. On the shift that began at 7 p.m. on Tuesday, December 3, Bertha was moving at a fast clip. Paul figures Bertha put up five or six concrete rings that evening, extending the tunnel by another about 30 to 36 feet. Sometime the following morning, on Wednesday, December 4, the operators began seeing 3- to 4-foot pieces of pipe and a bunch of smaller pieces showing up on the video feed of the conveyor belt. The workers werent particularly concerned.

When you hit something hard, usually there is a spike in the readings, says Paul. There was nothing. The machine had so much power [hitting the pipe], it didnt even show up on the screen. Indeed, from Wednesday morning through Thursday morning, Bertha continued to progress at a healthy pace, completing eight to nine rings in a 24-hour period.

Its not clear exactly when, but sometime early Thursday morning on December 5, operators monitoring the video feed began to see large boulders showing up on the conveyor belt. STP designers had specified that Bertha be built using a ribbon screw precisely because the massive screws hollow interior allowed it to pull rocks as much as 3 feet in diameter up onto the conveyor belt. At least once that morning, a rock emerged from the screw that was too large for one man to carry and that could damage the machinery in the interior of the TBM. The operators stopped Bertha so a hole could be drilled into the large rock and a screw attached so that a big ratchet arm with a lever hoist could be hooked to the screw to lift the boulder off the conveyor belt. The issue was quickly resolved and Bertha continued on its way.

Sometime that day, Jim and Paul arent sure exactly when, water started to collect in front of Bertha. One of the weaknesses of the ribbon screw that STP had specified when ordering Bertha from Hitachi Zosen was that it was less effective at maintaining pressure in the chamber while also allowing the removal of watery soil. Operators felt the need to use thousands of liters of a shaving-cream-like polymer conditioner to make the soil-and-water combination reach the toothpaste consistency required to pull it up from the ground. On Thursday, Bertha completed only three rings.

On Thursday night, Bertha, which normally moved at 20 to 40 millimeters a minute, was creeping along at 5 millimeters a minute. A Spanish operator was running the machine. The torque was at 5 percent, down from the normal 50 percent. That meant the machine wasnt scraping or cutting even though Bertha was operating at its maximum thrust of 100,000 kilonewtons.

We thought the machine might be bound up. Maybe pieces of pipe were caught in the teeth and in the chamber. Maybe the cutters were damaged, says Jim. It was clear there were problems. We had hit steel, we had hit boulders. Normal procedure would have been to stop the machine and see what was happening. But Dragados [the Spanish company that was the STP partner in charge of the dig] was making all the calls. Stopping the machine and checking the cutter head might have taken a week or two, precious time that could reduce the chance STP would earn the incentives built into the contract with WSDOT for completing the project early.

Working all night, the machine completed only two rings, extending the tunnel just 12 feet. On Friday morning, December 6, the machine had to stop multiple times to cool down. It was hardly making progress. Clearly, there were serious problems. What happened next was something neither Jim nor Paul witnessed, but which they heard about from a mechanic and one other worker present at the time.

An American operator was at the controls running the machine at the full thrust of 100,000 kilonewtons, when Juan Luis Magro, the STP executive from Dragados, removed the computer controls on the machine and ordered the American operator to increase the thrust as far as it would go. When the American operator refused, Magro ordered a Spanish operator to take over the controls.

A young Japanese engineer was present at the time. He was not happy with what the Spanish operators were doing, but the engineer believed his job was to offer technical advice. STP was the operator in charge, and Magro, the STP executive, was in command of the work site. If the STP operators insisted on pushing the machine beyond what its specs called for, and against the advice of the Japanese engineer, the engineer felt he was in no position to stop them.

The Spanish operator then proceeded to push the machines thrust to 325,000 kilonewtons, three times the maximum thrust Bertha was designed to handle.

It was crazy, an American mechanic told Jim later. He pulled out his iPhone and calculated that thrust was about 24 times greater than the thrust required to launch the space shuttle. The American mechanic told Jim that even the Spanish mechanic was furious about what his bosses were doing. The TBM began to overheat and the system automatically shut down to protect its critical seals and bearings. The Spaniards waited for the machine to cool down, then pushed it again. Its unclear how many times this process was repeated before they gave up. The two workers believe the overheating that occurred during this process is what damaged Bertha so severely.

Normal operating procedure would be to stop and check things out to see what was going on, says Jim. STP knew Bertha was still under warranty until it completed 200 rings. Since only 144 rings had been completed, the company wasnt worried about breaking it, figuring Hitachi Zosen would be required to fix the machine.

Jim figures that hitting the casing and then boulders likely damaged the cutters and they needed to be replaced. Paul agrees that things should have been checked but says the problem wasnt that the cutters were damaged, but that they were clogged. The cutter head had been designed to be thicker than past cutter heads to make it possible to replace cutters more easily during the dig, a specification STP had requested. The thickness of the cutter heads meant that the spaces through which the newly scraped soil was supposed to pass, were more easily clogged by pieces of steel mixed with polymer-filled soil.

Either way, if STP had stopped Bertha, replaced the broken cutters and cleared the goop that had collected in the cutter head, the machine might well have been able to continue on its way with a delay of a few weeks at most. Instead, the severe damage to the machine meant STP had to build a costly concrete pit to remove the cutter head and undertake substantial repairs that would result in a two-year delay.

The Spanish bosses were arrogant, says Jim. They overheated the machine. Ive worked with Germans and French. They are great. The Japanese from Hitachi were honorable, always helpful. But this bunch from Spain were something else.