Big Dig began with a critical decision
Novel technique may be behind troubles
The Big Dig was just a drawing-board sketch in 1988, a dream to replace Boston's traffic-choked elevated roadway with a wider, faster underground highway.
Contractor Bechtel/Parsons Brinckerhoff said it could be done and came up that year with a plan to make the subterranean roadway watertight: build a pair of massive slurry walls running from the street surface to the bedrock far below and dig out the dirt between them. Then construct a concrete tunnel box between the walls and cover the box with a membrane to seal it tight.
But by 1991, as the project moved from concept to construction, company engineers took a closer look at the narrow path for the new tunnel and confronted a sobering reality. To accommodate the inner tunnel box, they either had to reduce the size of the highway to six lanes, which would provide no improvement in traffic flow, compared with the elevated road. Or they would have to seize land under some of the most costly property in the United States, such as International Place or the Federal Reserve Bank.
So Bechtel/Parsons Brinckerhoff recommended a novel approach never before used for a highway tunnel in this country. Instead of building a tight box within the slurry walls, the contractors would use the rough, pock-faced slurry walls as the tunnel's only walls. That would save about 6 feet on either side, allowing for eight lanes of roadway. It would also save millions of dollars in land-taking and leave surrounding properties untouched.
State and federal officials agreed, work began, and for the next decade and a half that design decision would receive almost no attention amid the frenzy of construction that soon enveloped Boston. Even less heralded was the arcane, unprecedented challenge of adequately waterproofing 3 miles of tunnels with no interior walls.
It may prove to be the project's biggest challenge ahead: With this fall's revelation that hundreds of leaks riddle the brand new tunnels, engineers, taxpayers, and drivers alike are all confronting the results of that design decision 14 years ago.
A monthlong Globe review, based on hundreds of Big Dig memos, reports, contract documents, and dozens of interviews with officials involved with the project, found that the tunnels eventually built were a questionable substitute for what Bechtel/Parsons Brinckerhoff engineers initially envisioned and were especially susceptible to leaking if construction quality was less than perfect.
Lawmakers, lawyers, and federal and state transportation agencies are investigating two major kinds of leaks in the $14.6 billion project: small leaks where the rough slurry walls meet tunnel roof, and much larger breaches in the slurry walls themselves, the result of structural weaknesses caused by debris left in the walls by contractors.
Bechtel/Parsons Brinckerhoff officials, who declined to be interviewed for this story but replied via e-mail to a limited number of written questions, have insisted that both kinds of leaks are the result of poor quality work by construction contractors. Critics, however, say that the design was all but guaranteed to leak and that the fault for that lies with Bechtel/Parsons Brinckerhoff.
"They designed a tunnel knowing it would leak," said state Senator Steven A. Baddour, who cochairs the Legislature's Joint Transportation Committee. "They should have informed the state that this would have happened. Bechtel is held to a higher standard because of their expertise. That was their role."
Still, state officials went along with the approach, and even now some of them say that it was the only option that would work, given the land constrictions of downtown Boston.
"It was really a no-brainer," said Robert Albee, who in 1991 served as the state's director of construction services on the Big Dig. "It was not the kind of thing that turned into a big decision. Logic said this is the way you do it. And if you do it again today, you come to the same decision."
A radical proposal The original force behind the Big Dig was Frederick P. Salvucci, transportation secretary for Governor Michael S. Dukakis for much of the 1980s. Salvucci, a native Bostonian educated at the Massachusetts Institute of Technology, saw the project not only as a way to improve driving conditions for a gridlocked metropolis, but also as a way to heal the wounds of the first Central Artery project, which rammed the elevated highway through the West End and North End in the 1950s, with little regard for the area's residents.
As a result, Salvucci insisted that the Big Dig take no homes and leave the elevated highway open throughout construction. To do that, Salvucci embraced the idea of slurry walls, the stout concrete support structures perfected in Europe but relatively novel in the United States at the time.
The slurry walls would hold back earth and water from the excavation site, which would protect the foundations of the many nearby buildings. In addition, the walls would prop up the elevated artery while work proceeded beneath it. Unlike most projects at the time, which used slurry walls only during construction, the project Salvucci envisioned would use the slurry walls as a permanent part of the final tunnel structure.
Even so, Salvucci admits that his vision for the Big Dig was quite vague at the outset. Whether or not contractors would ultimately build a tunnel box within the slurry walls was not a detail he spent much time contemplating.
"That's a final engineering question," said Salvucci, who left office before the final design was chosen.
At the time, slurry walls, invented in the 1940s to hold back water and earth during underground construction, were generally removed or abandoned after a project was complete. Using them as permanent walls of a tunnel was a fairly radical idea. In fact, no major highway tunnels in the United States were built that way before the Big Dig.
As radical as the notion seemed, the federal government was keen on the idea as a way to save time and money. After all, governments in Europe and Japan were increasingly incorporating slurry walls as part of the permanent structure of tunnels. In 1979 the Federal Highway Administration held a symposium on the subject in Cambridge. The agency made it clear that it was hoping to take the idea nationwide, and what better place to begin than the Big Dig?
"We in the Federal Highway Administration contend that slurry wall technology is very underutilized in the United States today," said John Hooks, a senior Federal Highway Administration official, in his introduction to the symposium, which was later published. "We believe that there is a significant payoff to be realized by promoting the use of slurry walls on selected projects."
The design was used, in a more limited way, when the MBTA extended the Red and Orange lines in the 1980s. Bechtel Corp. of San Francisco and Parsons Brinckerhoff of New York played key roles in those projects, which used slurry walls as the sole walls of the T tunnels.
But the Big Dig took the concept to an entirely new level. It was to become the single largest use of slurry walls in the United States, with some of the walls extending more than 12 stories underground.
In April 1991, Bechtel/Parsons Brinckerhoff produced a thick report for the state evaluating seven proposed methods for building the Interstate 93 tunnels. Five of them envisioned a tunnel box within the slurry walls.
A sixth was an elaborate plan that would have used slurry walls alone, lodging the roof and floor of the tunnels into previously prepared notches in the slurry walls. But that required a high degree of precision, a difficult feat, given the enormous size of the Big Dig tunnels.
So Brian R. Brenner, the engineer who wrote the report, concluded that the only viable option was to go with the seventh alternative: slurry walls with no interior tunnel box and without precisely prepared sockets to connect the roof and the wall. The decision came down to space and money: To build eight lanes of traffic in a tunnel box would have required excavating tons more dirt, pouring more concrete and buying expensive nearby property.
Without the tunnel box, there would be enough room to build the roadway. "It is assumed that the right-of-way costs of taking adjacent structures is so prohibitive that it outweighs any advantage," Brenner wrote in the report.
But Brenner's recommendation to build the highway tunnel without inner walls flew in the face of an earlier report by other Bechtel/Parsons Brinckerhoff engineers on how best to waterproof the I-93 tunnel. In fact, the authors of that 1990 Big Dig waterproofing report never contemplated a tunnel without a separate set of neat, concrete tunnel walls inside the slurry walls.
The reason was plain: Without the separate inner walls, it would be difficult to connect the roof of the tunnel with the rough slurry walls in a way that could easily be made watertight. The connection between the 90-foot wide roof and 120-foot tall walls would be prone to expanding and contracting as temperatures fluctuated with the seasons. That meant that gaps and cracks were sure to open at the connection, leaving the tunnel prone to leaks from ground water above the tunnel if waterproofing measures at those junctures failed.
By comparison, a tunnel box would have involved a much smaller roof atop a far shorter inner wall that could be connected with a clean, flat seam, forming a far more watertight connection. Applying waterproofing material to that area would have been much easier and probably more effective.
In addition, Brenner highlighted a different potential problem in the design he was recommending: The outside of the massive slurry walls, the sides facing the soil, cannot be waterproofed. In other words, if the walls had flaws, ground water could force its way through and spill onto the roadway. That's exactly what happened Sept. 15, when water gushed into the tunnel.
The circumstances created by this daring design also made it imperative that whatever waterproofing materials could be applied inside the walls had to work.
Failures and fingerpointingWhile there was little that could be done about the outside of the walls, Bechtel/Parsons Brinckerhoff did take steps to make sure that inside the tunnel the connection between the roof and walls remained leak-free.
Bechtel/Parsons Brinckerhoff engineers directed contractors to cover the top of the tunnel roof with an expandable waterproofing membrane of liquid asphalt and spread it 1 foot up the slurry wall. In theory, such a method, along with other safeguards, would protect the roof-wall connection, even if the roof or walls moved slightly with temperature changes.
Despite those measures, as the tunnel began to take shape by 1997, Bechtel engineers had determined that the waterproofing of the wall and roof was not working as planned. In a report that year, the company said sloppy contractors were mostly to blame for the problem, saying they were installing the membranes incorrectly and failing to properly clean off the concrete on the roof, which left the waterproofing layer prone to puncture.
So Bechtel called for beefed-up inspections of contractors' work and ordered a new waterproofing approach on the roof, a sprayed-on membrane of polyurea, a newer product regarded as more reliable. The company also instructed contractors to install permanent hoses in the seam between the roof and walls that would allow maintenance workers in the future to shoot grout into any gaps and leaks that developed.
But subequent actions by Bechtel/Parsons Brinckerhoff suggest that it wasn't just the waterproofing material or the contractors that accounted for the problem. Rather, it was the company's approach to waterproofing a slurry wall-only tunnel.
A confidential report prepared in 2001 by Deloitte & Touche for the Massachusetts Turnpike Authority states that project managers had concluded that the initial plan to put waterproofing only 1 foot up the slurry walls above the roofs was inadequate.
After all, the ground water in the soil above the tunnels sat well above the 1-foot mark, allowing water to flow in behind any gaps or flaws in the waterproofing.
Bechtel then directed contractors to spray waterproofing all the way up the inside of the slurry walls and over the top, near street level, so ground water could not seep in from above. By then, much of the northbound tunnel had been built and covered with dirt and so was no longer accessible.
In fact, the year before, concern about the extent of leaks had already grown to the point that Bechtel/Parsons Brinckerhoff, the state, and federal officials had quietly established a leak task force to deal with the problem.
Correspondence from that task force shows that by the time the task force was up and running, it was probably too late to fix some of the leaks without resorting to more expensive and invasive measures.
One memo -- written on Dec. 12, 2000, by longtime Big Dig construction overseer Joseph Allegro, a state employee -- said that some contractors had never been instructed on how to properly prepare the slurry walls for waterproofing in the area where the roof and walls meet. Other contractors, who had been instructed in how to do it, simply didn't follow the instructions, Allegro wrote.
Finding fault As Allegro's memo underscores, relying on this unusual design for the project left little room for error in construction work. The Big Dig tunnel is entirely surrounded by soil sodden with salty ground water, which exerts tremendous pressure on the tunnel walls and roof deep underground. Any weakness or seam left unprotected would be vulnerable. Slurry wall specialists at the time cautioned that the approach risked leak problems.
Bechtel/Parsons Brinckerhoff officials declined to be interviewed for this story, but Brian Brenner, who wrote the key 1991 report recommending a tunnel design without inner walls, and Anthony Lancellotti, the firm's top design official, agreed to answer a limited number of written questions from the Globe.
Asked if, in retrospect, a concrete tunnel box would have been more watertight, Brenner and Lancellotti said that the question was moot, since they did not choose that path. However, they defended the design they chose, which has since become more common. If built properly, they said, it can be virtually leak-free.
They also said that comparing the Big Dig to other tunnels is unfair because of the unusual geographic constraints inherent in the downtown project.
"Use of slurry walls as part of the final structure have been widely used, in tunnels such as Harvard Square Station, the new Silver Line Transitway, the North Station project, and in most new garages in downtown Boston," Lancellotti and Brenner wrote.
"The waterproofing concept for the [roof-wall] joint is fairly standard. Other cut-and-cover tunnels cannot be exactly compared to the I-93 cut-and-cover tunnels because of the special design challenges associated with the width and depth of the I-93 tunnel, the need to build about half the tunnel under the low head room of the artery viaduct, also the need to underpin the existing artery viaduct, and the requirements for protection of historic adjacent structures," they wrote.
Baddour, who is one of many Massachusetts lawmakers urging that Bechtel/Parsons Brinckerhoff compensate the state for the problems, said he sees the management firm as the culpable party.
For the $2 billion the company is being paid to design and manage the Big Dig, its engineers should have come up with a plan for a tunnel that would both fit in downtown Boston and stay dry, he said.
"I don't buy the idea that it was either buy a billion dollars in artery real estate or build a leaky tunnel," Baddour said. "I think what happened here was a lack of creativity that created a disaster."
Sean P. Murphy of the Globe's staff contributed to this report. Raphael Lewis can be reached at firstname.lastname@example.org.