Engineers noted bolt slippage even after passed tests

Managers overseeing construction of the ill-fated ceiling in the Interstate 90 connector tunnel in 1999 were baffled to see ceiling bolts ``creeping out" of their holes even after they had passed a strength test, according to engineers' notes.

One bolt had slipped out of the concrete roof by more than a half-inch two months after passing its original strength test, and when workers retested the bolt, it ``began to pull out with almost no resistance."

Since the July 10 collapse of a section of the connector ceiling that killed a Jamaica Plain woman, inspectors have found similar slippage in 225 bolts in the connector tunnel.

In 1999, the failure of bolts after their initial testing led Big Dig managers to require testing of the bolts with more weight, but only after more than 80 percent of them had already been installed, according to a memo from the ceiling's builder, Modern Continental Construction Co. There are no records that workers ever went back and retested those bolts, which are held to the tunnel roof with epoxy.

Disclosure of the bolt problems -- including a Globe report yesterday about a 1999 memo in which a on-site safety inspector for Modern wrote to his bosses that he was ``gravely concerned" about the ceiling's safety -- has raised the possibility that Melina Del Valle died as a result of design and construction problems that managers were warned about seven years ago and yet did little to resolve.

Yesterday, one structural engineer said the tunnel ceiling should have been built with a wider margin of safety. More than 80 percent of the bolts apparently were tested up to a weight of 3,250 pounds, when the load they were supposed to carry was about 1,400 pounds. And the bolts passed if they held the test weight for only a few minutes. ``You're talking about a real, measured safety factor of 2.4, which is too low for this type of application, no question about it," said Stephen Buonopane, assistant professor of engineering at Bucknell University in Pennsylvania, referring to the ratio between the weight at which bolts failed tests compared to the weight they would normally bear. ``It should have been obvious to people."

He said the margin should have been at least 4 to 1, requiring tests of at least 5,000 pounds on each bolt.

Officials who oversaw the $14.6 billion Big Dig said yesterday they were relying on assurances from the company that supplied the ceiling bolts and epoxy, Powers Fasteners of New York, for their confidence in their strength. A senior Big Dig manager said Powers Fasteners had assured officials that the bolts could hold more than 20,000 pounds each, so they never expected the bolts to give way when tested with just a few thousand pounds. Officials at Modern Continental and Powers Fasteners declined to comment.

The Big Dig engineers' field notes from 1999 and 2000, stored in archives at the secretary of state's office, reflect concern among Bechtel/Parsons Brinckerhoff officials as an increasing number of epoxy bolts came loose when a weight of 3,250 pounds was suspended from them. In the first week of August, 1999, 14 out of 138 bolts, or 10 percent, failed their strength test and had to be replaced.

Of particular concern was the fact that at least five bolts seemingly held the weight, but then started to move later. One bolt passed the strength test on July 30, 1999, but moved more than a half-inch by Oct. 8. When workers pulled it out, they discovered most of the bolt was bare, suggesting that too little epoxy was used, and the epoxy still on the tip ``was brittle and easily crumbled," Bechtel/Parsons Brinckerhoff engineer Jim Barrett wrote, suggesting the two ingredients of the epoxy were not mixed properly. He also wrote that dust adhered to the epoxy, suggesting that the drill hole had been improperly cleaned.

``In spite of this combination of installation deficiencies, it appears the bolt was able to develop just enough strength to pass the pull-out test," Barrett wrote in notes dated Nov. 12, 1999.

Engineers from Powers Fasteners came to Boston for meetings twice that October, offering advice on how to get a stronger bond from the epoxy. The chemical can be stronger than concrete, but only if the bonded surfaces are clean, the epoxy is mixed just right, the drill hole is not too deep, and other conditions are met.

Eventually, Bechtel/Parsons Brinckerhoff called for testing all bolts with weights of 6,350 pounds. By then, work crews had already hung concrete ceiling panels from 1,000 of the epoxy bolts, and more testing would have required them to pull the panels back down. One construction official estimated that such a job would have cost at least $600,000.

No one has been able to find records that the ceiling bolts were ever retested, and construction officials involved in the Big Dig say it is highly unlikely that such expensive work would have been permitted, especially since Big Dig chief James J. Kerasiotes was zealously looking for ways to cut the Big Dig's bloated price tag. If someone had pitched a costly retesting program, said one construction industry specialist formerly involved in the Big Dig, ``the answer would not just be no. It would be hell no."

Scott Helman and Sean P. Murphy of the Globe staff contributed to this report. Scott Allen can be reached at allen@globe.com.