Boeing 787 chief project engineer Mike Sinnett says in hindsight he would have “challenged” the test assumptions in the evaluation and certification phase of the main and auxiliary power unit lithium-ion batteries for the 787.

“We had a battery supplier with a lot of experience, and with a test method that was state of the art,” Sinnett said during the opening session of a two-day NTSB investigative hearing into the failure of a Japan Airlines 787 auxiliary battery at Boston Logan International Airport after a flight from Narita International Airport.

“In retrospect, we may apply tighter test criteria or seek to understand test criteria a little more,” Sinnett added.

The NTSB during the hearing questioned FAA certification officials, Boeing and its subcontractors as part of its ongoing investigation into the Jan. 7 incident.

“There’s a lot of focus on returning to flight, but that’s not why we’re here,” said NTSB chairman Deborah Hersman during her opening remarks, referring to the FAA’s approval of Boeing’s fix for the battery problem on April 19. “We are here to understand why design failures occur with a leading manufacturer and a certification process that’s well respected throughout the international aviation community.”

Neither Boeing nor the NTSB have yet determined a root cause for the failure, which resulted in “thermal runaway” and propagation of failures to adjacent cells in the eight-cell auxiliary battery. Smoke and fire resulted, and one firefighter received minor injuries.

Boeing certified the battery under nine special conditions issued by the FAA in 2007.

Boeing subcontracts the power conversion system to Thales, which in turn subcontracted the lithium-ion batteries to GS-Yuasa. Sinnett says GS-Yuasa developed a test plan with Boeing’s approval that called for driving a nail through a cell to initiate a short circuit.

Sinnett says the resulting failure from that method “wasn’t as energetic as we’ve seen in service now.”

An alternative method Boeing investigated but did not use for certification involved a heating element that could be used to increase internal temperatures to produce a thermal runway, but Sinnett says the method could mask the results of the test by making it too conservative. “We heated batteries, but in no case did we get fire or flame from the battery,” he says. “We got thermal runaway inside the cell, but no cell-to-cell propagation.”

Boeing, however, did report propagation of thermal runaway between cells, as well as fire, when it performed overcharging tests, a key failure mode for lithium-ion batteries, but one that is controlled by the battery charging system. Sinnett says there is no evidence from the Boston incident that the battery was overcharged.

“The first thing that strikes me is that the overall system worked,” says Sinnett of the incident. “We didn’t have a catastrophic outcome. Clearly the event that happened was a serious event and we need to make sure it doesn’t happen again, but the added protection and layers of protection performed their function. From that perspective, it validates our process to a great degree.”