Transit center’s cracked beams not built to code
Official cites ‘failure of quality control in construction process,’ says defects preventable
The two beams that cracked at the Salesforce Transit Center were not built to code, and the cracks could have been prevented, the head of the agency that operates the center said Thursday.
For the first time, the agency has assigned blame for the cracked beams. Mark Zabaneh, executive director of the Transbay Joint Powers Authority, said companies contracted to construct the building — and multiple layers of oversight — failed to ensure that the building met code requirements.
“It’s a failure of quality control in the construction process,” Zabaneh said. “That’s why we are reviewing the rest of the building to make sure we don’t have any other incidents.”
The gleaming, $2.2 billion bus depot and eventual train station closed in September, just six weeks after opening, when workers installing ceiling panels happened to notice a large crack in a structural steel beam. A subsequent inspection revealed the second crack, both of which are in a section of the building that spans Fremont Street. Bus riders have been using the temporary terminal at Howard and Main streets since the discovery of the cracks.
Although repairs are already underway and expected to be completed by June, a date for reopening the center won’t be set until an independent peer review panel can determine whether there are similar construction problems in other areas of the building, said Christine Falvey, a spokeswoman for the authority.
There were two main factors that led to the beams failing: the quality of the steel used, and cracks that formed in the steel while the beams were being built, said Robert Vecchio, the president of New York- based LPI Inc., which performed a series of tests on samples from the steel to figure out why it cracked.
The American- made steel, which is 4 inches thick, was tough on the outside but soft in the center, he said, making it more susceptible to cracking.
The same steel was used in a different set of beams above First Street, but those beams didn’t crack even though they have the same design and bear the same amount of weight as beams above Fremont Street, he said. What differed between the two was the way the beams were built.
The First Street beams were constructed first, said Ron Alameida, director of project management for the authority, but quality control inspectors refused to accept them because they lacked weld access holes — a critical element used in welding to ensure the welds can be completed without any defects. So crews then cut the holes with torches after the beam was welded together, he said. The second time around, on Fremont Street, crews instead cut the access holes in first and then welded the beam together, Alameida said.
“Ironically the sequence on First Street was to address an issue of noncompliance,” he said. “The corrected work on Fremont Street is actually what saved (the beams on) First Street.”
That’s because on Fre- mont Street, when workers used a torch to cut the hole in the beam, tiny cracks formed. They should have first ground out the holes to make them smooth, eliminating any cracks. That was not done, Zabaneh said.
“It’s a concern from our part that (the grinding) was not done,” he said. “The code requires welding access holes to be ground, and these were not ground.”
It’s important, Zabaneh said, because those microcracks can grow larger, especially when more heat is applied to the steel.
And that is exactly what happened in this case, Vecchio said. After the access holes were finished, the beams were welded together. Welding takes place at very high temperatures, and that added heat caused the tiny cracks that already were present in the steel to grow larger, he said.
It’s unclear though, why multiple layers of oversight didn’t identify the lack of grinding as an issue, Zabaneh said. Herrick Corporation, a well-respected steel fabrication company based in Stockton, did the welding work and had its own quality control measures in place, he said. Then, Skanska USA, which was in charge of steel procurement, and Webcor- Obayashi, the general contractor, all performed their own quality control inspections.
The authority also hired Inspection Services Inc. to do a final inspection of the work. None of those layers of inspection identified the lack of grinding as an issue, Zabaneh said, even though it is visible to the eye and required in the code.
He added, however, that the project team and outside experts see this as an isolated incident.
Other experts noted that the lack of grinding and the way the beams were built were not the only factors that produced cracks in the beams. The center’s one- of-a-kind design, coupled with the materials used, also played a role, said Michael Engelhardt, a professor of structural engineering at the University of Texas and the chairman of the independent peer review panel.
“Certainly, the access holes played a role and the order in which they were placed played a role,” he said. “But it’s difficult to point at one item without looking at the whole package of items because they all happened together.”