Right now, as I type this, the world’s largest large-bore tunneling machine is sitting beneath the streets of my hometown in one of the most epic cases of bad project management in the history of project management.
The machine isn’t tunneling, and it definitely isn’t boring – it’s sitting idle, awaiting repairs and causing major headaches and stress to just about everyone involved. It’s part of an ambitious project to build a massive tunnel under downtown Seattle to replace a viaduct that would fall over if even the slightest earthquake struck.
The project is behind schedule, probably over budget, people are angry, and worst of all – no one really knows exactly how things are going to get back on track.
How did we get to this point, and more importantly, how can you avoid Bertha-sized failure on your projects? Join us in our first installment on learning from some of the biggest failures in project management history.
How Did We Get Here?
If you’re not familiar with Seattle’s tunnel woes, let’s back up a step. In 2001, the Seattle area was struck by the Nisqually Quake, a 6.8 magnitude trembler centered some 40 miles south of Seattle. It was one of the strongest earthquakes to hit Washington State in a very long time, and while it didn’t cause all that much damage, it did put a major hurting on the Alaskan Way Viaduct, an elevated freeway that runs through downtown. Built almost half a century prior on very loose intertidal soils, the Viaduct was already not in the best of shape. Carrying over 100,000 cars a day, it was and still is a major link in the region’s transportation network, and the Nisqually Quake rendered it barely safe for travel. One more tremor (heck, a strong gust of wind) and it was coming down.
Despite that seeming gun held to our collective heads, policymakers argued for years over the right replacement and funding. The options included rebuilding the viaduct, tunneling under the city, doing a shallower “cut and cover” tunnel or moving all that traffic onto surface streets, sort of like San Francisco’s Embarcadero. Public advisory votes were held and ignored. Finally, an agreement was reached to build a tunnel – a 54-foot diameter tunnel through soft glacial soils, the largest large-bore tunnel ever dug. To accomplish this, Hitachi was contracted to build a boring machine, which was named “Bertha,” after Seattle’s first and only female mayor, Bertha Knight Landes.
From the start, the project was beset by delays. The contract for construction of the tunnel had awarded some work to a union in contravention of an existing dockworkers’ union contract. The governor stepped in to craft a compromise, and Bertha got to digging. After digging just 1,000 feet, Bertha encountered resistance, which was first announced as a metal obstruction. Two months passed as crews worked to identify and remove the obstruction. When digging resumed, operating temperatures spiked, drilling was halted again, and it was eventually announced that the bearing seals on the cutting head had been damaged and that the grease behind the seals was contaminated with mud and silt. As of right now, the state agency in charge of the project and the contractor chosen to do the tunneling are working with Hiatchi to determine a repair plan, which may consist of digging a pit in front of the machine and lifting the entire 630-ton cutter head out of the ground.
No one knows when Bertha will resume digging or just how badly Bertha Knight Landes wants to haunt the living daylights out of the people who named the machine after her or how much all this is going to cost in overruns or who will pay for those overruns.
What Can We Learn from all This?
What can this massive mess teach you or me about project management? Plenty!
1. Prepare for contingencies.
When you’re starting a project, it’s worth conducting a thought exercise: what’s the worst that could happen? Brainstorm all of the things that can go wrong. The bigger the project and the higher the stakes, the longer that list should be. It shouldn’t be too hard to come up with. Once you’ve got the list, evaluate whether you can attach a probability to each of the contingencies. As Nate Silver has pointed out, a risk is something you can estimate the likelihood of. Uncertainty is something with unknown probability. Confuse these two concepts at your own peril! Once you have mapped out the points of failure, decide which ones you should have contingency plans against or even if it makes sense to proceed at all. In the cast of Bertha, probably someone should have looked at the risk of the machine getting stuck. You can actually calculate this, since there have been many large-bore tunneling projects in the history of infrastructure development. Then you have to account for uncertainty. This is the largest large-bore machine ever deployed. In the case of Bertha, it seems like the possibility of repairs should have been considered from the beginning, yet the contractor and the state are working with the builder of Bertha to determine how to fix the bearings. It’s as if this wasn’t worked out in advance, even though the possibility of needing to fix a large, complex machine while it was tunneling seems like an obvious contingency to be prepared for.
2. Always weigh the potential benefits and costs of project decisions, especially when the stakes are high.
During tunneling, the contractor responsible for operating Bertha met resistance and stopped the machine. Then they started it again and seemed to move forward. Then they ran it some more, until temperatures spiked. Then they stopped it again, paused for several weeks, started it again and finally found that the bearing seals were contaminated. What were they thinking when they kept running it the first time? Project managers should have realized they were in a marathon, not a sprint – any time gained in the short term would do nothing for them if it compromised the next year of drilling. In the presence of an anomaly – even a small one – the instant calculation should have been:
- What do we gain by running the machine again?
- What do we stand to lose?
- Which will I regret more – losing a day or two now or losing months later?
In that context, the choice would have been clear: you don’t simply rev the engine and see if it starts. Yet that’s what they did.
3. The greater the uncertainty, the greater the contingencies need to be.
If you’re doing something that has never been done before, you’re going to encounter things you didn’t know you didn’t know. “Oh, hey, it turns out that the unprecedented radius of the cutting head combined with the increased pressure caused by runoff and glacial silt compressing at 60 feet below sea level results in exponentially greater pressure being placed on the bearing seals when the entire mass is rotated, and our models for specifying the tolerances of that seal are inadequate at this scale.” (OK, I made that up, but it sounded plausible, right?) The more uncertainty, the greater your contingencies need to be either in time, money or planning.
In the case of this project, the contingency budget was $205 million on a total budget of $4.25 billion – in other words, less than 5%. As of right now, with the tunneling machine stuck and no real estimate on when it will be unstuck, that seems like a naively small number. Was there a lot of time put into contingency planning? Based on what’s been released so far, it appears the answer is no – the contractor and the state aren’t just trying to figure out what to do, they’re trying to figure out how they’re going to figure out what to do. That shows a shocking lack of foresight and a failure of basic project management acumen.
4. Communication is everything.
One of the theories as to why Bertha stopped tunneling was that it hit an obstruction. And an obstruction was found. Despite weeks of playful speculation wherein people’s imaginations ran wild about giant fossilized clams or abandoned logging equipment from Seattle’s skid road days, it turned out the obstruction was the metal casing of a surface well drilled to monitor groundwater. The embarrassing thing about this was that the state agency in charge of the project had put it there, left it there and knew it was there. However, they left it there on purpose and included its location in the RFP it released and explicitly stated that it was the bidder’s responsibility to remove it. Somehow, the winning contractor failed to connect the dots. In the end, that well casing may have had nothing to do with Bertha’s stoppage, but it’s a painful reminder that making sure the right information gets to the right place is waaaaaaay more than half the battle in making a project go smoothly.
So there you go – you are now better equipped to manage projects than the folks in charge of a $4.25billion mega-project.
How will it all unfold? Who knows, but the people of Washington are surprisingly used to this sort of thing. We built Galloping Gertie in 1940 just to see it fall into the Tacoma Narrows the same year. In 1979, the Hood Canal Bridge sank into Puget Sound in a windstorm. You would think we would have learned our lesson by then, but in 1990, we managed to flood the I-90 bridge and sink it into Lake Washington. We even managed to let a football stadium collapse while it was still being built. We’re like beavers in a flood plain: nature washes our work away, and we just stoically start re-building.
Hopefully, you can avoid that fate by doing things right the first time. As always, we wish you good fortune and successful collaboration!by