OVERBOARD!

The P&O Nedlloyd Mondriaan was steaming off the coast of the Netherlands with its decks stacked high with containers from the Far East on Feb. 9, 2006, when it was hit from astern with waves driven by winds of force 8 to 9 on the Beaufort scale. As the vessel rolled with the waves, 59 loaded containers tumbled overboard.

Ten days later, as the 7,500-TEU vessel was returning to the Far East with a cargo of empty containers, it ran into another storm in the Bay of Biscay, with headwinds of the same gale force. The ship lost another 50 containers. On the same day in the same storm in about the same location, the CMA CGM Otello, a vessel of approximately the same capacity but of a different design, lost 50 boxes.

These incidents were among the first of a mounting tally of containers lost overboard during the last two years. In 2006 and 2007, there were "significant" incidents where at least 36 ships lost a total of more than 1,600 boxes overboard. The full extent of the problem is unclear because there's no central repository for the data, and many ship lines understandably aren't eager to publicize lost containers.

Why are so many containers lost overboard? Is it the design of the new, larger ships? Or overweight containers? Or a sudden run of bad weather? Or the failure of the twistlocks that are supposed to hold the containers in place on deck? Or the cumulative motion of the waves, which causes parametric rolling by ships?

To find answers to these questions, several maritime organizations in the U.S. and Europe have undertaken studies. The technical services committee of the American Institute of Marine Underwriters has posted the results of its ongoing study on its Web site in a paper called "Ondeck Storage of Containers."

"Our concern is that there appears to be a trend of near-catastrophic losses of containers stowed on deck of container ships," said James Craig, president of the AIMU. "The purpose of the paper is to provide a better understanding of the risks associated with on deck storage of cargo and to further the public discussion as to the need for further regulatory requirements or industry safety initiatives."

In another study, Reederei Blue Star Line, the Hamburg-based vessel-chartering unit of P&O Nedlloyd, teamed up with the German classification society Germanischer Lloyd and the Technical University of Hamburg in 2006 to conduct a series of tests on container ships all over the world. The study group placed sensors in containers on different ships to measure the stresses they encountered during the last two winters. The tests were conducted on vessels chartered by Blue Star for P&O Nedlloyd, which was merged into Maersk Line at the start of 2007.

Although the tests haven't provided a definitive answer to the immediate cause of the losses, it is clear that they were due to a combination of factors, including, of course, weather, which is uncontrollable. "Unfortunately, on six ships in 2 years . . . we haven't found the actual cause in the end," said Lars Weissflog, a naval architect with Reederei Blue Star who is involved in the tests.

Harold Krill, managing director of Blue Star, said the problem with the P&O Nedlloyd Mondriaan may have been caused by the vessel's design. "We did discover that there are very high G-forces on the main deck levels in the stack behind the superstructure," he said. In both incidents on the P&O Nedlloyd Mondriaan and the one involving the CMA CGM Otello, the lost boxes had been stowed immediately aft of the ship's superstructure.

But Krill said he thinks a more likely cause of many of the incidents during the last two years is heavy or overweight containers loaded atop lighter containers. "There are more heavy containers being loaded on top of light containers because of the last-moment order when vessels with fixed schedules have a closing window of a few hours before departure when the last cargo can be booked," Krill said. "When you look into the last two years, we have more (capacity) available than cargo because there are many deliveries of new vessels that are bigger and bigger. Of course, the liner operators like to have as many containers as possible loaded onboard these vessels to offset the losses due to increasing fuel costs."

This would seem to point to a rather simple solution: Weigh all the containers before they are loaded on board so that their weight can be verified and they can be loaded properly. Many dockside container cranes are equipped with scales that can weigh containers. But this would cost money because longshore unions might demand additional personnel to weigh containers.

Krill said the possibility of weighing containers is being discussed at several ports, especially in England, where containers have collapsed aboard smaller vessels when there was just very slight movement because the containers on top are so heavy. "But you need another person to cross-check the weight with the manifest," he said. "This should not be a problem with the computer, but someone has to pay for it, and line operators don't like it."

Advocates of weighing containers say the cost could be offset in part by the revenue stream that would be generated by container lines billing shippers for loads that exceed declared weight.

Blue Star doesn't weigh containers before they are loaded on the vessels it charters on behalf of Maersk, but it does have an agreement with the charterers that it will not load containers that are more than two to four tons heavier than the container below it. Krill said Blue Star's onboard staff tracks the weight of the containers being loaded with a computer that displays the manifests with the weight of the cargo packed in a container.

That solution works on paper, but only on paper. If the weight of the cargo listed in the manifests is seriously underestimated, there is no way of finding out without weighing the container. The problem of overweight containers is being exacerbated in the U.S. by the rapid growth of containerized grain exports. "They are causing a lot of overweight problems," said Capt. James McNamara, president of the National Cargo Bureau, a nonprofit organization that exists to encourage safe stowage. "Many of the containers, when overloaded, are carrying weight they weren't designed for. They might not fail this voyage, but perhaps after they get discharged and come back in a different loading. The vibrations of going over the road or the railroad may cause metal fatigue and stress the welds. If the containers are weakened, they could give way when stacked on deck."

McNamara, too, said the solution to overweight containers is to weigh them before they are loaded. Some containers are weighed at U.S. ports, but only those that arrive by truck. "At present, I don't know of one terminal in the U.S. that weighs containers that arrive by rail," he said. That's why overweight shipments of containerized grain are causing so much trouble -- most are railed to their port of departure.

In the incidents involving the P&O Nedlloyd Mondriaan, Reederei Blue Star Line argued that the loss of containers was not caused by failure of the twistlocks that secured the containers, but that one of the containers on the lower tier had collapsed, causing the other containers to topple off.

"When we studied the loading plan, we discovered that there were heavy containers loaded on top and then together with the rolling of the vessel and the sea state and rolling, some of the containers on the containers on the first tier collapsed and everything on top of them started rolling," Krill said. "Then, of course, never mind what kind of twistlocks you have, everything started going."

The American Bureau of Shipping conducted a study in 2004 of ships' parametric rolling, which was the kind of rolling involved in the Mondriaan incident. The report carried the tongue-twisting title "ABS Guide for the Assessment of Parametric Roll Resonance in the Design of Container Carriers."

It found that some container ships are susceptible to parametric roll because of the fine form of the underwater hull geometry, combined with the full form above the waterline. It concluded that one way to reduce parametric roll is to install small anti-roll tanks to absorb the surplus energy collected during the parametric roll motions. The advantage of this method is that it wouldn't require intervention of a person or a computer.

Another method of prevention is to install fin stabilizers that move to counteract the roll motion of the ship. But the study admitted that, "The solution to the problem of parametric roll is not simple."

Peter Leach can be contacted at pleach@joc.com.