Designer proposed more bolts in Big Dig

Big Dig managers were so confident of the strength of the bolts that would hold up the heavy, concrete ceiling in the Interstate 90 connector tunnel that they successfully pressured the ceiling's designer to cut in half the number of bolts suspending each ceiling hanger, according to a 1998 memo obtained by the Globe.

``Unless I'm missing something, I don't see why [the ceiling designer] requires a 4 bolt anchor plate when a 2 bolt anchor plate would be sufficient," Robert Richard of Bechtel/Parsons Brinckerhoff, the joint venture that supervised the Big Dig, wrote to another joint venture official.

Then the bolts -- 20 of which popped out on July 10, causing a fatal ceiling collapse -- were safety-tested at a standard much lower than engineers at the time typically required in other construction projects. Bechtel/Parsons Brinckerhoff used a weaker safety test developed by an obscure Portugal-based group that was intended for bolts ``embedded in rock" -- not bolts in concrete roofs, like the ones in the connector tunnel.

``That's very disturbing to me," said Lee Mattis of California-based CEL Consulting, one of the nation's leading testers of the type of bolt used in the tunnel ceiling, called an epoxy bolt. He had never heard of the International Society for Rock Mechanics, the Portugal group whose ``pull-test" calls for placing 40 percent less weight on bolts than what Mattis says was the rule of thumb at the time.

Although there was no single standard in the 1990s for testing bolts in concrete, Mattis said Bechtel/Parsons Brinckerhoff should have used a heavier testing standard because epoxy bolts can fail without warning if the epoxy that holds the bolts in place is not applied properly. As a result, he said, construction projects rarely are designed with epoxy bolts to suspend heavy objects over people's heads.

``There's a reason for this: It's called gravity," said Mattis.

Bechtel/Parsons Brinckerhoff persisted in using the lower safety standard amid evidence that it was not tough enough to detect all faulty bolts. The standard was first used during construction of the Ted Williams Tunnel ceiling in 1994, and seven bolts that had passed the safety test came loose within weeks. Likewise, during the 1999 installation of the connector tunnel ceiling, five safety-tested bolts came loose shortly after the ceiling was hung from them.

Late in the project, managers increased the weight used for the safety test, but more than 80 percent of the epoxy bolts in the connector tunnel -- including the ones whose failure later caused the ceiling collapse that killed Milena Del Valle -- had already been installed. Bechtel/Parsons Brinckerhoff officials decided not to re-test those bolts at the higher weight.

Now, with a Suffolk County grand jury soon to hear testimony about possible crimes in the death of Del Valle , lawyers who specialize in complex death and injury cases say Bechtel/Parsons Brinckerhoff's actions could amount to criminal negligence.

``When they have actual knowledge that the system is not safe and they disregard that knowledge and continue to construct that way and create hazards for the motoring public, at that point, it's not simple negligence. To me, that is reckless conduct," said Anthony Tarricone , a Boston lawyer who frequently represents companies or victims in disasters such as plane crashes.

Bechtel/Parsons Brinckerhoff officials declined to comment for the record on their handling of the ceiling-bolt design and testing issues because the joint venture is the subject of a criminal probe. But speaking on the condition of anonymity, a consultant for the joint venture defended the safety testing, saying that the Society for Rock Mechanics is a reputable organization and that rock is similar enough to concrete for engineers to use the same testing standards. The consultant said the relatively low safety-test weight reflects engineering judgment, based on a belief that the weight on the ceiling bolts would be fairly constant.

Bechtel/Parsons Brinckerhoff acknowledges that it pushed the ceiling designer, a firm called Gannett Fleming, to use fewer ceiling bolts in the belief that properly installed epoxy bolts could hold several times the ceiling's weight. Gannett Fleming said in a statement that its engineers ultimately agreed that the ceiling would be safe even with half as many bolts holding it up.

``[From] everything I've heard from our lawyers, the genuine risk of criminal prosecution or criminal conviction is extraordinarily low," said the consultant to Bechtel/Parsons Brinckerhoff, who asked not to be named because he is not authorized to discuss the joint venture's legal strategy.

However, this fact remains: As the Globe reported last month, investigators said they found 225 loose epoxy bolts throughout the connector tunnel ceiling after it caved in, but none in the one area -- the high-occupancy vehicle lane -- where construction crews tested bolts with more weight than the Rock Mechanics' testing standard.

It is not clear who initially proposed using the Society for Rock Mechanics testing standard, which calls for bolts driven into rocks to be tested by pulling briefly on them with a weight 25 to 50 percent heavier than the weight they will normally carry. But by May 1992, Bechtel/Parsons Brinckerhoff, with the support of state and federal highway officials, had agreed that ceiling bolts in the Ted Williams and connector tunnels would be pull-tested at the low end of the range.

By contrast, Mattis of CEL Consulting said the rule of thumb was to test bolts at twice the weight they would carry under normal conditions. By 1995, before the connector ceiling was built, that was the standard for bolts used in California school construction and in hospital construction, for example, Mattis said. In addition, the Nuclear Regulatory Commission required testing at twice the planned weight for bolts in nuclear reactors, a field in which one part of the Big Dig management group, Bechtel Corporation, was a leading contractor.

Bolts in the connector ceiling would be tested with just 3,250 pounds, compared with 5,200 pounds if Bechtel/Parsons Brinckerhoff had followed Mattis's rule of thumb.

Though Bechtel/Parsons Brinckerhoff defends the decision to use the Rock Mechanics' safety test, for much of the project, Big Dig managers erroneously described the source of the testing standard in key documents. In guidance documents sent to the companies that built the ceilings, the Rock Mechanics' group is not mentioned.

Instead, Bechtel/Parsons Brinckerhoff wrote that the testing standard was based on work by a highly respected standard-setting group for the construction industry, the American Society for Testing and Materials. However, the ASTM standard listed in the guidance documents does not provide a safety testing standard for bolts. Bechtel/Parsons Brinckerhoff officials could not explain the discrepancy.

Bolt experts say the bolt-testing standard used on the Big Dig is particularly suspect because an epoxy bolt is much trickier to install than a conventional bolt, with workers having to follow detailed steps to make sure the epoxy attains its full strength.

Bechtel/Parsons Brinckerhoff should have used more weight than it did in the bolt-safety tests, according to David Fowler , an engineer at the University of Texas at Austin who teaches courses in the proper use of epoxy bolts.

Gannett Fleming initially objected to the use of epoxy bolts for the connector tunnel ceiling, and when Bechtel insisted, the designers created a blueprint that called for the ceiling panels to be suspended from groups of four epoxy bolts attached to a hanger.

But, by the late 1990s, Bechtel/Parsons Brinckerhoff placed great faith in the new bolts, and Richard of Bechtel/Parsons Brinckerhoff believed that Gannett Fleming was being too conservative.

He said in his June 23, 1998, memo that two epoxy bolts could safely hold more than 14,000 pounds, which was far more than the estimated weight held by each hanger.

Richard told Gary Baxter of Bechtel/Parsons Brinckerhoff that Gannett Fleming should either cut the number of bolts in half or submit detailed calculations ``proving that 2 anchor bolts do not work," according to the memo.

Gannett Fleming officials ultimately agreed to reduce the number of bolts, calculating that the ceiling would be safe ``assuming proper installation and quality of the product materials," according to a Gannett Fleming statement.

But that turned out not to be a good assumption. Workers had many difficulties installing the epoxy bolts in the connector tunnel roof, sometimes failing to mix the two ingredients in the epoxy correctly, other times not putting enough of it in bolt holes, according to field engineers' reports. Workers also used a drill bit not recommended for use with epoxy bolts. As a result, dozens of bolts pulled out when they were given the pull-test.

There was a ``larger concern," as was noted by a Bechtel/Parsons Brinckerhoff engineer on the Ted Williams Tunnel ceiling project. Sabine McDuffie warned in a memo to her boss, design chief Anthony Lancellotti , that some bolts were passing the pull-test when they were, in fact, defectively installed.

In the connector, bolts also came loose though they had passed the Rock Mechanics' safety test.

If the tunnel ceiling were being built today, Mattis of CEL Consulting said Bechtel/Parsons Brinckerhoff would have had to closely inspect bolt installation to prevent mistakes in the first place. In addition, the Concrete Anchor Manufacturers Association issued guidelines in 2002 that would have required a pull-test of 10,000 to 15,500 pounds instead of the 3,250 used in most of the tunnel. Mattis said bolt experts now understand that the lower-weight pull-tests only catch grossly inadequate bolts.

But, even in 1999 when construction of the connector ceiling began, Mattis said there was plenty of information available that could have cautioned Big Dig officials about the risks of using epoxy bolts and the need for a tougher safety program.

``It's just very unfortunate that someone had to die," he said.

Scott Allen can be reached at allen@globe.com. Sean P. Murphy can be reached at smurphy@globe.com.