right_division Green SCM Distribution
Bookmark us
SCDigest Logo

Focus: Global Supply Chain and Logistics

Our Weekly Feature Article on Topics Related to Global SupplyChain Logistics

From SCDigest's On-Target e-Magazine

- Oct. 23, 2012 -


Global Logistics News: MIT Research Puts Some Real Data Behind Ocean Shipping Variability

Lots of Complaints, but Little Data to Date; Delays Come at Ports, not Over the Water, Caplice and Kalkanci Say at CSCMP


SCDigest Editorial Staff


There is a lot of variability in global ocean shipping times, right?

Well, that's certainly what a high percentage of the corporate sponsors of MIT's Center for Transportation and Logistics were telling director Chris Caplice and others at MIT, Caplice recently said during a presentation at the CSCMP conference in Atlanta.

SCDigest Says:


At the Port of LA, the performance varies dramatically, with the best performer on average processing containers two and a half times faster than the worse, and everyone else in between.

What Do You Say?
Click Here to Send Us Your Comments
Click Here to See Reader Feedback

As Caplice note, however, the problem is that "the evidence was all anecdotal" - almost none of the shippers and importers had hard data to support their complaints about inconsistent delivery times.

So, Caplice and colleague Basak Kalkanci decided to see if they could actually quantify what was going on, and the two were able to get data from a handful of shippers, freight forwarders and others. With that data, they were to perform some analysis to begin the journey of better quantifying what is really happening in the container shipper market.

Caplice started off by noting that one issue right upfront is that there is far from a universal definition of how "reliability" in the ocean shipping sector is defined.

He said that some companies focus primarily on whether the carrier did what it was supposed to do, or what he called "credibility" (e.g., didn't bump containers, made all the stops they were supposed to, etc.).

The other perspective is what he called "schedule consistency," which relates to such issues as whether the containers arrive on schedule, or (from another angle) how much variance there is for a carrier around their average or median performance (i.e., how tight is the bell curve?).

A related challenge to understanding what performance really is, both Caplice and some audience members noted, is the fact that while syndicated date from researchers such as Drewry Shipping Consultants provides accurate data on ship departure date and arrivals versus published schedules, that data is only port-to-port, and does not tell you the full lead times from container arrival at the origin port or availability at the destination. The focus of the MIT research and analysis was at the specific container level, not the ship level.

And that is quite important because, as Kalkanci detailed, it is at the outbound and inbound port operations where in fact most of the delay and variation is occurring. More on that below.

Caplice also noted that there are too often disconnects between the procurement staff negotiating the contracts with carriers and the operations team that must execute them. Too often those contracts are "simply thrown over the wall" to operations, he said, and there may be factors that make it very difficult for carriers to meet the service levels in the contract - but by then it is too late.

Another issue is that the penalties and bonuses for failing to meet or exceeding service schedules are rarely applied, Kalkanci said, which again brings into relief the disconnect between the contracting processes, where these incentives are negotiated, and the actual execution of those contracts.

"Token penalties don't motivate carriers to make real changes to the operations," Kalkanci added.

She then showed data illustrating that port-to-port delivery times were pretty consistent, whereas dwell times in origin and destination ports were much more variable. That can seen in the chart below, which illustrates a common measure of variability, the coefficient of variation, for different elements of the total ocean journey. As can be seen, the measure is markedly lower for the on-ocean part of the full trip than for port operations versus pre and post-ocean shipping steps. (The coefficient of variation is calculated by dividing the standard deviation of a data set by its average.)

Ocean Shippping Variance by Step in Total Move, by Coefficient of Variation



Source: Caplice and Kalkanci, MIT

The idea port-to-port reliability is a good measure of an ocean carrier's overall performance is a myth, Kalkanci said.

(Global Supply Chain Article Continued Below)



Wide Variations at the Port of Los Angeles

That data showing that most of the variation in ocean shipping comes from port operations not on the water variability was supported by another very interesting analysis Caplice and Kalkanci presented. That analysis was performed on a data set from a freight forwarder that tracked container handling times across ocean carriers at the Port of Los Angeles.

The chart below shows relative variations in the average handling times at the port across 18 carriers and more than 7000 containers shipments.


Source: Caplice and Kalkanci, MIT



As can be seen, the performance varies dramatically, with the best performer (carrier A) on average processing containers two and a half times faster than the worse (carrier S), and everyone else in between.

Reducing Variation Has Greater Impact that Reducing Average Lead Time

Finally, Caplice and Kalkanci presented data showing that in general, shippers would be better off working to reduce variation in total lead times than they would reducing average lead times in ocean shipping.

As shown in the chart below, they estimate that companies can decrease inventory levels by 9% on average by reducing shipping time variation by three days, versus a 7.7% reduction in average shipping times. (Those numbers come in relation to smartly using the new estimated lead times, of course – not everyone would see these gains.)


Percent Potential Decrease in Inventory for Reduction in Lead Times



Source: Caplice and Kalkanci, MIT



They also said, however, that reducing average transit time pays off more if the transit time of the lane is short and consistent (e.g., China to West), while reducing variability pays off more if the transit time of the lane is short, but variable.

Caplice also noted at the end of the presentation that the total impact on shippers from the widespread practice by major carriers today of "slow steaming" – or moving much slower than a ship is capable of to reduce fuel costs and effective capacity – is not well understood. He said he and Kalkanci will be looking at this topic and doing more research on the variability topic in coming months.

Gene Tyndall, executive VP at consulting firm Tompkins International, has been advising shippers to focus more on variance than averages for many years.

"The combination of demand volatility and global supply chain complexity continue to plague companies that need to optimize inventories and deployments," Tyndall told SCDigest. "The issues are not in port-to-port travel times, but are rather in the variability of times for end-to-end, including the key processes of loading, steaming, and off-loading, and customs clearance.

He added: "When these vary, all supply chain plans and synchronizations get thrown off, requiring safety stocks at each point in the chain. Predictability is a critical success factor in plans and synchronizations, thus in achieving capital efficiency."

Any company interested in potentially confidentially sharing some of their data for MIT to analyze for you and the industry can talk it over with Caplice by sending him an email. We encourage you to help support this on-going research.

Do we need better understanding of what is really happening with ocean shipping reliability? Why is it so hard to get the data for analysis? Let us know your thoughts at the Feedback section below.

Recent Feedback


No Feedback on this article yet