ANCHORAGE- When the Exxon Valdez pulled away from the Alyeska oil terminal on the evening of March 23, its crew had just put in a grueling 22 hours, loading 53 million gallons of North Slope crude oil into the ship's cargo tanks.
Whether fatigue from that long night and day helped put the 987 foot tanker on Bligh Reef three hours later is under study by the National Transportation Safety Board.
But testimony taken by the board and other records made public since the spill make it clear how how hard the crew was worked during the ship's stop in Valdez last spring.
The load on the three mates who serve immediately under the captain was particularly intense.
During the 24 hours that Alyeska allots each tanker to take on its cargo, one of the mates had to be on the bridge at all times, while the other two oversaw the ship's docking and mooring, the loading of its cargo, and its undocking and sailing.
So heavy was their workload, according to the records, that none of them appears to have had enough rest to legally take a bridge watch when the ship pulled out of port shortly after 9 p.m., March 23.
Yet one of them Third Mate Gregory Cousins was in charge of the bridge when the Exxon Valdez plowed into Bligh Reef shortly after midnight on the morning of March 24.
Experts on sleep deprivation say that work schedules and conditions like those that prevailed on the Exxon Valdez could have played a part in the crash.
"We've looked at some of the literature on the Exxon Valdez and of course the main point is that the timing of the accident was consistent with a sleep related or fatigue related error," said Dr. Merrill Mitler, scientific director of the Scripps Clinic Sleep Disorders Center near San Diego.
Cousins' day, according to the records, went like this.
He had gone to bed at about 1 a.m., March 23, after finishing the 8 p.m.midnight watch on the bridge. Just more than six hours later, he was awakened for his next watch, which began at 8 a.m.
He finished the watch at noon, went to the ship's engine room to take some measurements, and returned to his quarters at about 1:30 p.m. After 30 minutes' paperwork, he told the safety board, he took a "catnap."
That catnap, according to the available records, is the only real rest Cousins had during the 16 hours before the tanker ran aground, and it didn't last long.
Chief Mate James Kunkel, Cousins' boss, saw him eating dinner at 4:30 p.m., meaning the nap couldn't have lasted much over two hours. Kunkel told the safety board he was surprised to see Cousins up at that time.
By 5 p.m., at any rate, Cousins was back at work, relieving Kunkel so that he could get some supper.
After that, Cousins helped finish loading the cargo, helped with the undocking, then took up his watch on the bridge just before 10 p.m.
There Capt. Joseph Hazelwood, using radar to pierce the darkness and drizzle of the spring night, was feeling his way down a narrow slot between islands and reefs on the ship's left side, and a field of icebergs from Columbia Glacier on the right.
Shortly before midnight, Hazelwood turned the bridge over to Cousins and went below.
Minutes later, the third mate apparently missed a turn and the Exxon Valdez hit Bligh Reef, pouring 11 million gallons of its cargo into Prince William Sound. It had been more than 16 hours since Cousins reported for his first watch of the day.
There aren't many restrictions on how hard the crew of a merchant ship can be worked, but such restrictions as there are come from the federal government. The most specific is a law that prohibits a shipowner or operator from permitting an officer to take over the bridge of a ship leaving port unless he's been off duty for at least six hours out of the preceding 12.
Based on his own testimony and that of his shipmates, Cousins probably hadn't had much more than three hours rest out of the previous 12 when the Exxon Valdez left port March 23, an apparent violation of the federal law.
When William Deppe, an executive of Exxon Shipping Company, was asked at the safety board hearings how the company made sure the law was followed on its ships, he answered that it didn't.
"We don't have any program to give six hours of rest to any deck officer before we get under way," Deppe said.
Scot Tiernan, a commander in the Marine Safety Division at the Coast Guard's Alaska headquarters in Juneau, said he didn't know if the rest requirement specifically was under investigation. But, he said, "If there was any violation of federal law or regulation, I'm quite sure that's being looked at."
The state of Alaska, in a July 17 filing with the safety board on the causes and circumstances of the accident, didn't appear to have much doubt it was violated.
"It is difficult for the State to understand how ESC (Exxon Shipping Company) can dispatch its tanker fleet with so little regard for the law or for the alertness and readiness of its crews," the state wrote.
Exxon, in its own filing with the safety board, didn't address the question of possible violations of federal law, merely contending "there is no credible evidence that fatigue contributed to the cause of the grounding."
Although Cousins was still on the bridge when the ship hit the reef shortly after midnight, he wasn't supposed to be. Cousins was supposed to have been relieved by the second mate, Lloyd LeCain, at 11:45 p.m., about 20 minutes before the crash and in plenty of time to have turned the ship away from the reef.
But Cousins told the safety board he stayed on watch so LeCain could sleep.
LeCain, according to the NTSB records, worked the midnight4 a.m. watch on the morning of March 23, went to bed, got up at 9 a.m., worked the noon4 p.m. watch, then worked at various cargoloading and undocking chores until the ship sailed. At 10 p.m., he went to bed, and was still in it when the ship hit the reef.
Thus LeCain, if he had been awakened on schedule at 11:20 p.m. to take his midwatch, would have had approximately 11|2 hours rest out of the previous 12, also far less than the six required by federal law.
First Mate James Kunkel had a tough schedule, too. Like Cousins and LeCain, he was interviewed by the Coast Guard just after the crash, and testified at the safety board hearing seven weeks later.
Kunkel's scheduled watches were from 48 a.m. and from 48 p.m. While the ship was in port, Kunkel stood watches, oversaw the cargo loading, and was on the bridge at the engine controls when the Valdez pulled out in midevening, March 23.
Cousins relieved him a few minutes before 10 p.m., and Kunkel went to his room, took a shower, and went to sleep. Like LeCain, he was in bed when the Valdez sailed into Bligh Reef.
"I had been up for most of the time during the load," Kunkel told the safety board, "and I finally was able to say to myself, "Well, the job is done, I can now go to bed and go into a deep sleep,' which is what I tried to do."
Kunkel told the safety board that even under normal sea conditions with no cargo to be loaded or unloaded he routinely worked 12 to 14hour days.
Taking the 48 watches, Kunkel said, allowed him to carry out his other duties in the intervening eight hours.
"I would normally try to work, or I would have to work, usually anywheres from four to six hours during the . . . day period," he said.
"Are you frequently up 24hour stretches?" asked Eric Sager, a safety board staff member.
"Yes," Kunkel replied. "However, very seldom have I been on my feet for a full 24hour period. There is always an hour or an hourandahalf that I can sit in my chair."
The problem with people who get too little sleep, or get it in quick snatches at odd times, is not that they can't perform the tasks they normally perform on the job, the sleep experts say. In fact, laboratory tests show that sleepy people, when kept busy, are almost as effective as wellrested people.
"You can't really tell a sleepdeprived basketball player when he's playing," said Mitler. "You couldn't even tell a sleepdeprived air traffic controller if you looked at him for 30 seconds in a crisis situation."
The problem with sleepy people is what happens when they're not busy, when there's no crisis, when the job is routine. They may miss, or may misinterpret, the early signs of a problem, allowing a minor mistake or malfunction to balloon into a catastrophe.
Or, they may simply fall asleep and not realize it later.
"The ability to maintain constant vigilance is what's dropping off," said Donald Tepas, an industrial psychologist at the University of Connecticut. "People have what are referred to as microsleeps. They sleep for seconds or perhaps minutes."
Cousins told the safety board he was wellrested as he worked on the Exxon Valdez in the hours leading up to the crash, but he was unable to explain what happened during a critical five minutes just before the accident.
Cousins was ordered by Hazelwood to start turning away from Bligh Reef and back toward the standard tanker lanes when the ship passed Busby Island, a little more than two miles north of the reef. That happened at 11:55 p.m. and Cousins said he ordered the turn within a minute of taking the fix.
But the course recorder on the ship showed the turn didn't start until five to six minutes later, at 12:01 a.m.
At the safety board hearings, Cousins was asked to explain the gap.
"I've studied the course recorder for a long time, and I really don't have an adequate answer for that," he said.
Then he was asked if he could simply have lost track of time for a few minutes.
"No. No," Cousins insisted.
Tepas, the Connecticut sleep expert, said the problem of nodding off or missing cues while on duty is especially acute for watchstanders. Their jobs tend to be monotonous and repetitious at best. And they often have to work at night, or when they're tired, or as in the case of Cousins both.
"You have similar problems in the rail industry, and with cross country truck drivers," Tepas said. "A lot of those truck accidents are from truck drivers falling asleep at the wheel."
Another factor is technology, Tepas said. Automation has made most watchstander type jobs easier, and therefore even more monotonous.
"A lot of things people had to do before automation that kind of kept them awake aren't there any more," Tepas said. "Instead of being machine operators, they're now machine watchers."
And the work ethic on oil tankers could make the problem worse, Tepas speculates.
"Especially when you're talking about people who are kind of macho, which is probably true of a crew like this, it's not unreasonable to suspect either that they were not able to detect how sleep deprivation was affecting their performance, or they were unwilling to admit it," Tepas said.
Like Cousins, Kunkel denied ever being too tired to do his work. But he told the safety board that fatigue is part of the job, part of getting ahead in the merchant marine.
"This is just normal to me," Kunkel said. "If you want to be the chief mate, this is how you will work. If you want to be a second mate or a third mate, this is how you will work. If you want to be a captain, this is what is expected of you."