The main engineer on the tow ship for Shell's Kulluk drilling rig testified on Thursday that its engines failed because of a problem with the diesel fuel that materialized in the form of "slime."
The Coast Guard took samples of the fuel for testing but Carl Broekhuis, chief engineer on the Aiviq, told a Coast Guard panel investigating the Kulluk's Dec. 31 grounding that he never saw the results.
The substance, whatever it was, caused every fuel injector for all four of Aiviq's main engines to fail.
"This wasn't just slime. There was something else on it," Broekhuis said, describing a picture of a goo-coated fuel filter that is part of the Coast Guard evidence. "It was something that was unexplained."
Broekhuis testified on Day 4 of what is expected to be a 10-day Coast Guard hearing, and the focus was largely on the Aiviq, a new 365-foot vessel built by a Louisiana company, Edison Chouest Offshore, to Royal Dutch Shell's specifications for drilling in Alaska.
His account provided the most direct explanation to date that the fuel was the cause of the engine problems, but investigators also asked a series of questions about fuel vents that became submerged during an earlier tow.
The Coast Guard testing of the fuel is continuing, Cmdr. Joshua McTaggart, who is leading the investigation, said during a break in the hearing Thursday.
Also on Thursday, Bobby Newill, an Aiviq third mate with Coast Guard credentials as a master of towing -- the vessel's "tow dude," according to his attorney -- testified about how surprised he was when the primary tow gear failed the morning of Dec. 27 while he was on watch.
The Aiviq was towing the Kulluk on what was supposed to be a weeks-long journey from Dutch Harbor to a Seattle-area shipyard for off-season maintenance when the initial tow line broke. A series of emergency tow lines were rigged between the Kulluk and various vessels that came to help, but none held.
Newill called the weather "moderate" on Dec. 27 and said everything seemed normal before a shackle failed and was lost at sea.
What about, McTaggart asked, the numerous alarms that morning from a Rolls Royce monitoring system that indicated too much tension on the tow line?
"Do you recall receiving these alarms while you were standing watch?" McTaggart asked.
No, Newill answered. "I've been asked this question before, and I still say my answer is I did not receive a tension alarm for the tow drum."
Tension alarms had been going off frequently for a different piece of equipment, the starboard anchor drum, but that was due to a faulty sensor, Newill said. That equipment wasn't even being used during the transit. Repair parts have been ordered from Norway, he said.
The alarms emit an ear-piercing sound and must be addressed with four steps, not like a snooze button on an alarm clock, he said.
"I attended to every alarm," Newill said. "You can't brush it off."
Newill offered up his own piece of evidence, a video he said he shot on his phone earlier on the morning of the grounding to share with his wife later.
"If it pleases you guys, I did make videos from time to time," he said.
McTaggart conferred with others on the panel and decided to allow the video, which had a time mark from Newill's phone of 9:47 a.m. on Dec. 27, less than two hours before the tow line broke for the first time.
Newill narrated as his minute-long video played. Investigators also played a five-minute video from the closed-circuit television system of the Aiviq that captured the moment of the tow failure.
The "tensile strength overload" alarm was set to sound if the pull topped 300 tons -- half the tow wire rating of 600 tons, Newill said.
His video showed the monitor flashing readings that showed the tow line tension fluctuating dramatically, from a low of 20-odd tons to a high of 227 tons. He contended the average was 100 tons, well within the equipment's strength.
McTaggart said that still was a lot of fluctuation.
The other video didn't show the console but provided a view of the Kulluk under tow. The seas were as others had described -- giant swells but not fast hard waves. At the end, the Aiviq seems to ride way up, and then the line snaps and whips back.
That same day, the No. 2 engine was the first to go, just before 11 p.m. By then an emergency tow line had been hooked up. The other three engines went down early Dec. 28.
Broekhuis told the Coast Guard panel he was able to keep the ship steady with a powerful thruster, kind of like an outboard motor. Using spare fuel injectors on board, his crew brought the No. 1 engine back up by 6 a.m. on the 28th. Edison Chouest flew many more injectors to Kodiak in owner Gary Chouest's jet, and the Coast Guard delivered the parts early on Dec. 29.
Broekhuis testified that he suspected a fuel additive -- put in before Aiviq fueled up in Dutch Harbor -- caused the engine problems. Todd Case, the Kulluk rig manager for Noble Drilling Corp., told him he had experienced a similar injector issue, and the additive was the culprit. Dutch Harbor fishermen told him of fuel problems, too.
Broekhuis and his attorney, Ken Schoolcraft, declined to say where the Aiviq got its fuel.
After a replacement injector failed, Broekhuis switched to a fuel tank that he was confident had good fuel, and no more injectors failed.
Still the Aiviq burned almost all of the fuel it had on board. He didn't explain directly how the engines functioned with fuel from slimed tanks. The ship has many systems to ensure clean fuel including filters, centrifuges and settling tanks, he said.
The fuel vents were flagged as a concern after water got that high during an earlier trip. But he tested the fuel in numerous spots and never found much water, Broekhuis said. He did find a jelly-like substance.
Edison Chouest decided to raise the height of the vents to diminish the possibility of water getting in, he said.
It cleaned all the Aiviq tanks and discarded the remaining fuel in Seward.
It's now reworking the procedure manual for the vessel.
The Coast Guard investigation could lead to recommendations for systemic safety improvements or actions against the licenses of individual crew members.