Time for a rethink

Time for a rethink Drewry says carriers' on-time performance regressed in late 2009, but how to truly measure schedule reliability?

By Eric Johnson

      A recent survey of ocean carriers' on-time performance by Drewry Shipping Consultants revealed that schedule reliability got worse, not better, toward the end of 2009 ' a bitter development for shippers who enjoyed low freight rates last year, but are being stung by significant rate hikes in 2010 without an apparent elevation in industry-wide service quality.
      In theory, carriers' on-time performance was supposed to be dramatically better in the second half of 2009. Two factors were working for the carriers:
      ' Services were being slowed down to save on fuel consumption and emissions, allowing carriers a buffer in their schedules.
      ' Ports were less congested due to the downturn in demand for containerized goods.
      Yet, according to research by London-based Drewry, on-time performance in the last quarter of 2009 significantly worsened from the first half of the year. From a high of nearly 68 percent in the second quarter of 2009, carriers were on time only 53 percent of the time in the last quarter.
      There are many potential explanations for the deterioration in on-time arrivals at the end of 2009. Carriers, as they began slowing their loops, added ports to the rotations on those loops. More stops equals more chances for delays. Inclement fall and winter weather could be contributing factors, as could potential disruptions from the addition or suspension of new services.
      Drewry's measurement is but a snapshot of a very large, nuanced, and complicated picture, and there are those who would argue about what exactly constitutes 'on-time.'
      For its survey, Schedule Reliability Insight, Drewry tracks the on-time performance of about 1,600 ships in 10 ports worldwide and defines 'on-time' as arriving on the scheduled day or one day early. Yet some would argue it's more accurate to judge schedule reliability by hours, not days.
      For instance, here was a scenario put to American Shipper recently by a former shipping line executive: Ship A is scheduled to arrive at 11 a.m. on May 1, but actually arrives at 11 p.m. on the same day. Ship B, meanwhile, is scheduled to arrive at 11 p.m. on May 1, but actually arrives at 1 a.m. on May 2.
      So Ship A is 12 hours late while Ship B is only 2 hours late. However, because Ship A arrived on the correct date, it is considered to be on time, while Ship B is deemed to be late. What's more, Ship A has arrived so late that it has blown through the shifts of two longshore gangs, while Ship B might still be worked by the gang that was meant to unload it.
      There are countless other hypothetical examples that cloud the picture.
      'We look two weeks in advance at the carriers' schedules and cross-check that with what information we get from the ports,' said Simon Heaney, editor of Drewry's Freight Shipper Insight and Schedule Reliability Insight. 'It's not 100 percent perfect. We're definitely a broad-gauge measure.'
      One issue is that the information available is often not very specific. Swiss carrier Mediterranean Shipping Co., for example, only provides the day that its vessels are due to arrive, not a specific time. So it's impossible to track data any more closely than a carrier provides.
      'We can only work with the data that we can extract from the carriers' schedule,' Heaney said.
      While some may quibble with the methodology, what you can call Drewry's reliability index is unique. There's nowhere else to turn, at this point, for an objective industry-wide look at how late or early ships are arriving. And it's an increasingly important aspect of container shipping.
      'It's an appalling picture when 60 percent is the best in our industry when it comes to on-time reliability,' said Eivind Kolding chief executive of Maersk Line, at a major shipping conference in March. 'We need to get to the same state as the parcel industry, which has 90 percent-plus on-time.'
      Kolding showed Drewry statistics that often cite the Danish line as the best, or among the best, in terms of on-time performance. But, in terms of improving reliability going forward, history is not on his side.
      Drewry's historical on-time data shows that, with the exception of the second quarter of 2009, industry-wide on-time performance has stayed within a band between 45 percent and 60 percent since the end of 2005. And the second quarter 2009 performance can likely be explained by the fact that demand was so low and services so scaled back that ships couldn't help but be on time more often.
      It's not an enviable track record.

Slow Steaming. Drewry theorized, as did many analysts, that slow steaming would improve the on-time performance of carriers. Indeed, carriers themselves partly justified the move to slow down services by saying it would add a margin of error to their schedules that often didn't exist before. On an express service, there's little chance to make up time lost at a congestion-filled port, or at sea due to a weather diversion.
      'One of the touted benefits of slow steaming was supposed to be that it would give them a buffer for their schedules,' Heaney said. 'But carriers are so cost-focused. Even if they are faced with a situation where they are behind schedule, they aren't interested in speeding up to get on time.'
   There's no motivation for carriers to speed up and incur extra fuel costs for schedule reliability, said Francis Phillips of American Shipper's ocean capacity research affiliate ComPair Data.
   'When ships get out of place in their loops there is a significant cost in quickly putting them back into regular same day of the week order,' he said. 'Honest schedules nearly always look irregular for a few weeks ahead, only gradually settling back into a precise, predicted, pro forma pattern. While revenues are down, why should the lines constantly spend a lot making their schedules look neat and tidy by speeding up ships to get them quickly back into place?'
      One shipper, though, said he expects better reliability in exchange for slow steaming.
      'There was initially some concern about slow steaming, but it's not necessarily a big issue to go slow,' said Jeff Siewert, director of international logistics for Home Depot, at a ports forecast event in Long Beach, Calif., in late March. 'But going slow should improve reliability. The cargo should get there when it's supposed to get there. Slow steaming should help carriers hit their on-time targets, because that helps us plan the rest of our supply chain.'
      Part of the problem, Heaney said, is that shippers don't hold carriers accountable for delays. And carriers have never been able to gain any consistent financial advantage through faster transit times.
      'Even at peak times, carriers have never charged a premium for fast service,' he said.

Service Changes. One explanation for the poor performance seen in the last three months of 2010 was service upheaval. On a regular basis, loops were being withdrawn or rearranged, vessels pulled, idled or switched to other services.
      'The nature of the service restructuring, with vessels going from different regions, out of layup, it didn't help,' Heaney said.
   But those type of service rearrangements are normal for carriers, said Lamont Petersen, vice president of market for Hyundai Merchant Marine (Americas).
      'Some schedules were temporarily affected by movement between services and/or dry dock but this is not unique to 2009,' Petersen said. 'Carriers traditionally make seasonal changes to vessel deployments twice annually to coincide with the import peak season.'
      Meanwhile, Heaney said it's entirely possible that the traditional factors that caused poor schedule reliability in the past could crop up again this year.
      'Congestion may well become a factor in the next couple of quarters in terms of meeting berthing windows,' Heaney said. 'Looking forward, if demand picks up, unless they have an incentive to tackle this, it will inevitably get worse.'
      That, of course, depends on a continued recovery in volumes. Right now, recovery looks more solid in Europe than in North America ' though economists believe a surge in underlying demand for containerized goods is still a ways off in both regions. And since the Asia/Europe lane is far longer than the transpacific, late arrivals could be a bigger factor in that trade. Ships surveyed by Drewry the past 15 quarters were an average of 1.17 days late between Asia and Europe and 0.81 days late between Asia and North America. Ironically, the shortest lane, the transatlantic, has suffered the worst, with ships an average of 1.37 days late.
      'Most lines automatically allow an extra day for long transits in the winter and remove it in the summer,' Phillips said. 'Making allowance for weather at sea is not to be underestimated, even with big ships, because good seamanship is an art where safety comes first, not like running a railroad on dry land. Slowing down for a storm has nothing to do with saving fuel.'

Port Calls To Blame? Another potential factor for the poor reliability figures is added port calls. As strings have slowed, carriers have realized that more ports can be added to rotations without affecting schedule integrity, at least theoretically. But added port calls create more potential pitfalls. Ships can be held in port by bad weather, or issues with loading or unloading.
      'The whole line is that schedules are kept to weekly fixed sailings,' Heaney said. 'But the lines are simply not putting out reasonable schedules. It's unrealistic. And the more ports you put in a rotation, the more chance for delays.'
      But an executive with Singapore-based carrier APL said slow steaming will help improve on-time performance.
      'If, as we believe, slow steaming is here for the long term, both carriers and their customers will need to change how product is flowed through supply chains,' said Jim Dwyer, vice president global network operations for APL. 'It's true that carriers are adding days to schedules, which means more time in port and on the ocean. But this can also offer a buffer to boost reliability. For example, where necessary we can speed vessels up to ensure they are on time. Also, during summer peak periods, popular loading ports can get congested and many are prone to weather delays during the off-peak. By having more contingency capacity on the ocean we can accommodate these scenarios and still perform reliably.'
      As for the complications of adding more ports, Dwyer said there are benefits to such actions as well.
      'If you put additional ships into a service you can call at additional ports, which can also reduce feeder costs,' he said. 'Increased schedule flexibility will also allow ships to berth during different windows ' docking on weekdays or during daylight hours. A key point is that our services don't really operate independently. Rather, they're part of an interconnected global network ' with most loops intersecting one another at some point. For the most part, adding more ships helps to ensure service loops operate at optimum efficiency.'
      Carriers cannot afford to overlook adding a port call on slowed services, Petersen said.
      'Typically slow steaming means adding another vessel to a string, and with the addition of a new vessel there is too much additional time in the previous schedule even after slowing down ' thus the addition of ports to improve the efficiency of the vessel deployment,' he said. 'The addition of port calls in itself should not affect reliability unless that call is particularly subject to inclement weather.'
      But Phillips pointed out there are backlogs in ports ' even with little congestion ' that can throw carriers too far off their schedule to recover.
      'Port delays can often be out of a line's control if a ship in front is late off its berth,' he said. 'Again weather can be a factor. Backlogged terminals must sometimes decide where to concentrate mobile equipment ' working either shipside, or shoreside on trucks and trains. The scheduled departure may not be the actual departure.'

Reliability For Shippers. APL considers itself a market leader in on-time performance and it's a selling point the line stresses. The line finished fifth-best in the latest Drewry survey.
      'Customer surveys consistently place the highest priority on service predictability and schedule integrity,' Dwyer said. 'APL has always been among the most reliable carriers ' this is central to our global service proposition.
      'We compile a detailed service reliability scorecard. Each week we measure the on-time arrivals and departures of APL vessels and those of our partner carriers. We adhere to very strict criteria ' a vessel must arrive and depart within four hours of the published schedule.'
   That, of course, underscores the challenge that those tasked with measuring 'on-time' across the industry face. If one line measures by day and the other by four-hour increments, an apples-to-apples comparison is impossible.
      For example, Hyundai (APL's partner in the New World Alliance) deems anything later than six hours to be late.
      'For purposes of measuring 'internal' on-time performance, if a vessel is delayed by six hours or more it is considered late,' Petersen said. 'The targeted string performance is 90 percent or better on-time. Within (the New World Alliance), on-time for commercial purposes is considered to be arriving on the same day as the pro forma arrival. Arriving early counts as on-time but frequently the vessel is constrained by berth availability, so it will usually slow down to arrive within several hours of pro forma rather than arrive early and wait at anchor or sit idle alongside.'
      Dwyer said APL's focus on schedule integrity is necessary for it to deliver on the promises it makes as part of its time-definite services.
      'Meticulous planning and control over our operations and assets enables us to meet our reliability standards,' he said. 'This has also enabled us to pioneer the time-definite services ' OceanGuaranteed and APLGuaranteed ' that we offer with our sister business, APL Logistics. These services have a money-back guarantee for failure to deliver on the agreed day. So far, on-time performance has been near perfect.'
      He also said that on-time performance wasn't affected much by changes to APL's service network.

Eivind Kolding Chief executive officer, Maersk Line
'It's an appalling picture when 60 percent is the best in the industry when it comes to on-time reliability. We need to get to the same state as the parcel industry, which has 90 percent-plus-on-time.'

      'We reduced deployed capacity and made significant changes to our global service network to adjust to significantly lower demand conditions,' Dwyer said. 'But we also added ships to ensure capacity was where customers needed it and that services continued to run on time.'
      Aside from Drewry's survey, the trade publication Logistics Management queries shippers each year about their opinions of ocean carriers, with a specific mention of on-time performance.
      LM's survey provides an even more inexact measurement of who is actually on-time, since it asks shippers for their opinions of carriers' on-time performance, then rolls that score into a larger overall score to subjectively determine the highest quality carriers.
      While the LM measurement may not be mathematically watertight, it does provide an interesting counterpoint to Drewry's more objective assessment. In essence, it allows the industry to weigh the perception of who is on time with the reality.
      For instance, in August 2009 (its most recent survey), LM found that among global carriers, 'K' Line, Hamburg S'd and OOCL had the best on-time performances. According to Drewry, those lines finished 17th, ninth, and 12th, respectively in the last quarter of 2009. That's quite a disconnect.
      What it mostly proves, as Heaney suggested, is that no measurement is perfect. It's tough to compare carriers within alliances (with similar networks, pooled vessels and operational similarities); tougher still to compare carriers across the industry. And so a true measure of industry-wide schedule reliability may remain elusive.