INTRODUCTION

Globalisation of the world economy and the associated profound changes in worldwide manufacturing and distribution processes to a large degree have been made possible by the invention and development of container shipping. Container shipping has changed the way we transport goods around the world. By enabling easier access to the exchange of goods, it has opened up new global markets for export and import. Containerisation has provided the mechanism to expand to international markets without sacrifice to the quality of distribution and without incurring a major penalty in high freight costs.

Containerisation has also facilitated 'just in time' (JIT) production through its improved schedule reliability, low costs, high security and faster transport times. Container shipping lines have developed or at least aim to develop liner services networks characterised by low operating costs, high frequencies, fast transit times, and both tight and reliable voyage schedules. With a growing complexity in global transport networks, managing the factor time in the design and operation of liner services is not an easy task. Port congestion is just one of the many factors that can disrupt schedules and as such negatively affect schedule reliability. Time is money. Lost minutes result in lost dollars for shipping lines and their clients. In the competitive market place faced by shipping lines every dollar counts.

This contribution deals with the time factor in liner services design and operation. The first section discusses the design parameters in liner service design as well as the wide range of liner service patterns carriers offer to meet customers requirements in terms of frequency, transit time and price. The second part of the paper offers an analysis of transit times and schedule reliability. Sources of delays are identified and evaluated. Furthermore, the range of measures aimed at reducing schedule unreliability is discussed. The concepts developed and issues discussed will be applied to loops operational on the East Asia–Europe trade, in particular between ports in China, Japan and Korea and the load centres in the Hamburg–Le Havre range in northern Europe.

DESIGNING A LINER SERVICE SCHEDULE

Design parameters

Schedule design is most often referred to as a strategic planning problem for shipping lines (Fagerholt, 2004). When designing a liner service, carriers are considering primarily the market to be served. There are three key inter-related decisions for service planners:

• The service frequency. Carriers will try to have at least a weekly service. In doing so, they make a trade-off between frequency and volume on the trunk lines: smaller unit capacities allow more frequent services and as such meet shippers' demand for lower transit times, while larger units will allow operators to benefit from economies of vessel size.
• Fleet size, vessel size and fleet mix. The optimal vessel size depends on cargo availability, shippers' needs for transit time or other service elements and the choices made with respect to the other two key variables. As economies of vessel size are more significant on longer distances, the biggest vessels are deployed on the longest routes (see eg Cullinane et al, 1999; Lim, 1998). The fleet size and mix also constitute important strategic planning aspects in schedule design. Carriers have to secure enough vessels to guarantee the desired frequency. Large size differences among the vessels operating within the same schedule could decrease operational homogeneity. Shipping lines typically strive for a well-balanced fleet mix within the individual loops they operate. There are, however, operational and financial barriers to a shockwave increase in vessel size, so the fleet mix might not always be so homogeneous. Upgrading the vessel size on a specific route can take several years and demands huge phased investments.
• The number of port calls. Limiting the number of port calls will shorten round voyage time and increases the number of round trips per year, thereby minimising the number of vessels required for that specific liner service. However, fewer ports of call mean poorer access to more cargo catchment areas. Adding port calls can generate additional revenue if the additional costs from added calls are more than offset by revenue growth.

Carriers design the networks they find convenient to offer, but at the same time they have to provide the services their customers want in terms of frequency, direct accessibility and transit times. This tension between routing and demand is important. The network planners may direct flows along paths that are optimal for the system, with the lowest cost for the entire network being achieved by indirect routing via hubs and the amalgamation of flows. However, the more efficient the network from the carrier's point of view, the less convenient that network could be for shippers' needs. Shippers could avoid the indirect routes, opening the possibilities for other shipping lines to fill gaps in the market.

A wide range of liner service patterns

When observing recent developments in liner shipping, there has been a clear drive to reduce the number of ships operated in relation to overall revenue. The productivity has been improved by using faster and larger ships, but of comparable importance in this respect are the devising of new operational patterns and cooperation between shipping lines.

Shipping lines today have a wide range of patterns at their disposal, all of proven merit in particular circumstances. Triangle services, pendulum services, butterfly services, conveyor belt services and other forms of varying complexities are mixed with simple end-to-end services, and adapted for both mainhaul and relay services to create a network best fitting a carrier's requirements (for terminology on types of liner services see eg Notteboom, 2004). This growing complexity in liner service networks is in line with the findings of Robinson (1998). In referring to the Asian hub/feeder restructuring, he argues that a system of hub ports as main articulation points between mainline and feeder nets is being replaced by a hierarchical set of networks reflecting differing cost/efficiency levels in the market. High-order service networks will have fewer ports of call and bigger vessels than lower order networks. Increasing volumes as such can lead to an increasing segmentation in liner service networks and a hierarchy in hubs. Gilman (1999) argued the networks operated by large vessels continue to be based on end-to-end services. Hub-and-spoke systems are just a part of the overall scene.

In the struggle to maximise carryings, relaying has played an important role, particularly in filling surplus east–west slots. Nevertheless, only a handful of lines have built relay networks that effectively involve the full integration of trade routes. Maersk Sealand is a prime example. The post-Panamax ships deployed on its pendulum services not only provide slots on the Far East and Europe/North America, but also act as a conveyor belt between a series of controlled hubs – notably Algeciras, Salalah and Tanjung Pelepas. Virtually all the carrier's cargo to/from West Africa moves through Algeciras, from which weekly loops radiate. Most of these loops are 'double loop' or mini-pendulums. The main difference between Maersk Sealand's relaying and that of many other carriers is the close integration of all parts. Different services dovetail to provide smooth connections, and operations at the main hubs are effectively under its control. The only other liner operator to have made serious steps in this direction is MSC, which has several firmly established relay services, and launched several mini-pendulums (eg on the west Australia/Singapore/Thailand route). Mini-pendulums not only give extra direct services, but also offer a safety valve in case of delays.

Both Maersk Sealand and MSC have the scale to generate enough cargo to use the ships they employ. For the strategic alliances and groupings (Grand Alliance, New World Alliance, etc), such a strategy is unlikely given the different priorities of the members. Few dedicated relay services have been started under joint banners, and integrated operations in the Maersk Sealand mode are unlikely.

Line-bundling networks or loops

Notwithstanding the increased significance of relay-type of networks, pendulum services and other alternative network configurations, most liner services on the main shipping routes are of the line-bundling type (Notteboom, 2004). A line bundling loop is conceived as a set of x roundtrips of y vessels each with a similar calling pattern in terms of the order of port calls and time intervals between two consecutive port calls. By the overlay of these x roundtrips, shipping lines can offer a desired calling frequency in each of the ports of call of the loop. As mentioned earlier, intercontinental liner services typically have a calling frequency of one vessel per week.

In the remainder of this paper, we will mainly focus on the time factor in traditional loops of the line-bundling type. The concepts developed and issues discussed will be applied to loops operational on the East Asia–Europe trade, in particular between ports in China, Japan and Korea and the load centres in the Hamburg–Le Havre range in northern Europe. The latter port range includes the ports of Antwerp, Hamburg, Rotterdam, Le Havre, Bremen and Zeebrugge. The loops considered are listed in Table 1. More complex liner service configurations do not fall within the scope of the paper. The loops considered on an average call at around 10 ports of which six are in Asia and 4 are in Europe. To secure a weekly call, carriers typically deploy eight vessels per loop, with each vessel having a total round voyage time of around 50–55 days. Average vessel size now amounts to more than 5,600 TEU (post-Panamax class). In recent months, units of more than 8,000 TEU have been introduced on the trade and many more will enter the market soon.

Table 1 - Selected loops on the East Asia Europe route (January 2005).

Full table

The service patterns on the East Asia–Europe route have substantially changed in recent years as a result of China's economic boom. Chinese ports have recorded tremendous growth in container throughput. The lure of potential high returns has led international terminal operators such as Hutchison, PSA Corp and APM Terminals to pump money into Chinese ports, bringing infrastructure and service in line with global standards. Rising volumes and upgraded infrastructure in Chinese ports make it more attractive for carriers to increase the number of direct calls, rather than rotating containers out by feeder to regional hubs (Meyrick, 2004). This rescheduling of liner services to and from China has particularly affected the Korean container port system with major ports Busan, Gwangyang and Incheon (Yap et al, 2003). Only a few years ago, Busan thrived as the regional hub for Chinese export cargo. At present, Busan and neighboring Korean ports are struggling to keep Chinese transhipment cargo flows. Whereas Chinese cargo in many main ports around the world has recorded double-digit growth figures in 2004, volumes of Chinese cargo handled at Busan are down 2% over 2003.

TRANSIT TIMES AND SCHEDULE RELIABILITY

The container transport system is structured by time-tight schedules. Shipping lines have developed a strong focus on designing liner services with short transit times, combined with a high degree of schedule reliability. In view of delivering an impeccable service to their customers, shipping lines are keen to meet the timings as announced in the official (published) schedules. Delays not only decrease the reliability of the liner service, but can also incur logistics costs to the customer as a consequence of additional inventory costs and in some cases additional production costs (eg a production stop due to a late delivery of materials). It goes without saying that delays also incur costs on shipping lines in the form of additional operating costs, linked, for example, to unproductive vessel time and the rescheduling of vessels. Sailing speeds can provide some slack, but this comes at a cost (see further in this paper).

Depending on the market segment, the relative importance of each of these quality dimensions may differ. For example, a low number of late arrivals might be a decisive competitive factor in a mature market where competing lines are not otherwise differentiated and transit times are comparable. In other situations, being first to market may take precedence over schedule reliability. For the first liner service in a new liner service design concept, transit times as such may be relatively more important than the number of late arrivals. While the relative importance of each quality dimension may change in proportion to the others, the absolute importance of each is high in the liner shipping business.

Transit times

In a narrow approach, the transit time can be defined as the number of sailing days on a port-to-port basis. In a broader logistics chain approach, the transit time is the total time on a door-to-door basis, so including dwell times at terminals and time needed for pre- and end-haul to the port of loading and from the port of discharge. As the focus in this paper is on the maritime segment, that is, liner services on the Europe–East Asia trade, we will primarily use the narrow definition (port-to-port).

The average transit times for the loops introduced earlier in this paper are depicted in Table 2. It can be observed that transit times do not differ substantially. The time to sail from the last port of call in Asia (mostly Singapore or Port Kelang) to the first port of call in Europe is relatively stable at around 16 days (including Suez transit). The ports of call in East Asia are typically spread over a very long stretch of coastline (eg from Dalian in the north to Port Kelang near the Straits of Malacca). Together with a higher number of port calls in Asia, this explains the difference in intra-regional time in East Asia versus Europe. Figure 1 depicts the decomposed transit time for a typical loop on the East Asia–Europe trade. Port time in this particular case absorbs about 21% of total time.

Figure 1.

Decomposition of the scheduled transit time for the CEU service of Cosco.
Source: Based on data COSCO

Full figure and legend (61K)

Table 2 - Composed transit times in days for 24 loops on the East Asia–Europe link.

Full table

A crucial element in port-to-port transit time is the order of ports of call in the loop or string. If the container's port of loading is the last port of call on the maritime line-bundling service and the port of discharge the first port of call then transit time is minimised (see eg Lago et al, 2001). In practice, shipping lines' decisions on the order of ports of call is influenced by many commercial and operational determinants, including the cargo generating effect of the port (outgoing cargo), the distribution of container origins and destinations over the hinterland, the berth allocation profile of a port, the nautical access, etc. A port regularly featuring as last port of loading or first port of discharge in a liner service schedule in principle has more chance of achieving a higher deep sea call efficiency ratio compared to rival ports who are stuck in the middle of the loop. Table 3 depicts the deep sea call efficiency ratios for the main load centres in the Le Havre–Hamburg range. The high deep sea call efficiency ratios of upstream ports like Antwerp and Hamburg may be explained by the need for a high traffic volume to warrant the out-of-pocket costs and time costs of sailing up the river (ie diversion distance).

Table 3 - Deepsea call efficiency ratio for selected European ports (TEU handled as a percentage of two-way capacity of vessel called at port).

Full table

Shipping lines, terminal operators and also shippers are well aware of the huge variations in transit times incurred by the position of a port in the port calling order. In maximising service levels to the customer, many shipping lines have developed several shipping options, ranging from a lowest price – lowest transit time option to fast transit times at a higher door-to-door price (ie higher price mostly caused by fast inland transit). For example, when transporting goods from East Asia to southern Germany, shipping lines could leverage on the transit time advantages of Mediterranean ports compared to northern ports. Given appropriate rail infrastructure, transit time would cut down to 17 days versus 22 days (Figure 2). Potential benefits from this scheme depend on the reliability of the Mediterranean option versus the northern option, the sailing frequency to the Mediterranean port of call versus the northern ports, and the proportion of the cargo that is sufficiently time sensitive to warrant such an undertaking.

Figure 2.

Total transit times (Asian port of loading to European inland destination).
Source: Based on data provided by carriers and Eurogate

Full figure and legend (105K)

Offering short transit times is a competitive factor in liner shipping, in particular when the goods involved are time sensitive. Typical examples are perishable goods and consumer goods with a short life cycle or elevated economic/technical depreciation (fashion, computers, etc). An extra day at sea creates opportunity costs linked to fixed capital and could lower the economic value of the goods concerned. The average value of containerised goods differs substantially among trade routes (see example in Table 4). A delay of one day incurred by a container load with a value 40,000 typically results in the following costs: (1) opportunity costs (3%–4% per year)=3–4.5 per day and (2) economic depreciation (typically 10%–30% per year for consumer products)=10–30 per day. One day of delay with a post-Panamax vessel carrying 4,000 full TEUs from the Far East to Belgium thus implies extra costs on the goods of at least 57,000 (at 3% opportunity cost and 10% depreciation), which is much higher than the charter rate per day for a post-Panamax container ship (at present around USD 40,000 per day). The costs of delays imposed on the cargo could even be more elevated in case major disruptions take place in a production line due to late delivery of containerised raw or semi-finished products. As such, delays in the maritime leg of the chain have an impact on the total door-to-door transit time, which is built into the supply chain inventory model. The shipping and transport markets have developed pricing mechanisms to anticipate expected delays (eg congestion surcharges) or to price delays ex-post (demurrage fees).

Table 4 - Average value in of goods per fully-laden TEU for different trade routes (import and export containers in relation to Belgium).

Full table

Delays

Delays and time losses in vessel operations within a loop can have many causes. As depicted in Figure 3, the causes of delays can be classified into four groups: terminal operations, port access, maritime passages and chance.

Figure 3.

Functional elements in a line-bundling network or loop

Full figure and legend (83K)

Port/terminal congestion or unexpected waiting times before berthing or before starting loading/discharging

Rising port volumes and capacity constraints in many ports around the world mean that berth availability on arrival in a port is not always guaranteed when the allocated time slots in the ports have been missed. In some instances, port congestion can completely disrupt liner service schedules. For instance, congestion in southern Californian terminals during 2004 saw fully loaded vessels backed up for up to 10 days waiting to berth and unload. The cargo delays that have occurred in southern California, especially in the twin ports of Los Angeles and Long Beach, are changing the way trade moves across the Pacific to the US. More shipping lines are either calling at more northern ports (Seattle, Tacoma or Vancouver) or bypassing the US west coast and instead going via the Panama Canal to the US east coast ports. An example of the former is Seattle. With its own gateway terminal in Seattle, APL was able to ease some of the congestion by running a loop of vessels up to Seattle in order to keep cargo flows moving for its key customers. Rising congestion has made some shippers think differently about when to ship cargo, which in itself results in changes in ordering and shipping patterns. Also in the East Asia–Europe trade, port and terminal congestion remains a critical issue, forcing shippers and shipping lines to change business strategies.

Port/terminal productivity below expectations (loading/discharging)

Contracts between shipping lines and independent terminal operators generally contain specifications on the required minimum quayside productivity. At present, required terminal productivities of 120 TEU per ship per hour are no longer an exception, leading to high expectations both in terms of gantry crane availability and the speed of quayside operations. In scheduling a loop and calculating transit times, shipping lines take into account the expected distribution in terminal performance. They also take into account the flexibility of terminal operators in dealing with 'emergency' situations.

Port congestion and port productivity are incentives for shipping lines to secure capacity in key ports in their service schedules. In Europe for instance, shipping lines have gradually entered the market via the development of dedicated terminals at major load centres (Table 5). Dedicated terminals are even more widespread in Asia and North America. Drewry Shipping Consultants (2003) collected throughput figures for terminals in which carriers have a non-minority shareholding: Evergreen handled 5.7 million TEU worldwide on its terminals in 2002, Cosco 4.7 million TEU, Hanjin 4.7, APL 4.3, NYK Line 3.5 (including 1.3 million TEU at its subsidiary Ceres Terminals), OOCL 3, NOL 2.5, K-Line 2.2, MSC 2.2, Yang Ming 1.3 and Hyundai 1.1 million TEU. Container shipping lines approach terminal management in a different way: they seek control over berths while other 'pure' terminal operating companies manage multi-user facilities. Many of these liner terminals offer stevedoring services to third carriers as well, thereby creating some hybrid form in between pure dedicated facilities and independently operated multi-user facilities.

Table 5 - Some examples of shipping lines' direct interest in European terminals (Summer 2005).

Full table

Access channels

Access channels often constitute bottlenecks in the global maritime transportation system. Disruptions in port access come in various forms, ranging from unexpected waiting times due to irregularities in pilotage or towage services (eg low availability of pilots or tug boats) to unexpected waiting times caused by delays at sea locks or the morphology of the access channel in terms of tidal windows.

The port of Antwerp is a prime example of a port subject to lock operations and a tidal access channel of 80 km (the river Scheldt). When calling at Antwerp, a container vessel can sail closer to the economic markets in the hinterland. However, post-Panamax vessels are limited by the available draft, and hence the state of the tide, and so are restricted to certain tidal windows when sailing up and down the river. The largest container vessels need to take into account some tidal window restrictions, although these have been substantially widened via a series of deepening programs. Concerns about maritime accessibility also exist in Hamburg and Bremerhaven to name but a few. Tidal windows in maritime access channels complicate the design of a liner service schedule and could lead to changes in the order of port calls. On the Europe–East Asia trade, draft limitations and tidal windows are becoming important design variables, as many Chinese container ports and a number of load centres in the Le Havre–Hamburg range are increasingly facing difficulties in accommodating the latest generation of container vessels at all times.

Maritime passages

Maritime passages such as the Suez Canal and Panama Canal play a significant role in carrier schedules. The Suez Canal forms a 163-km ship canal in Egypt between Port Said on the Mediterranean and Suez on the Red Sea. The canal allows ships with up to 15 m of draft to pass, and improvements are planned to increase this to 22 m by 2010 to allow supertanker passage. There is one shipping lane with several passing areas. The passage takes between 11 and 16 h. Owing to the limited width of the canal, ship convoys are formed on either side of the canal. When a container vessel arrives late at the Canal, it misses the convoy of which it was planned to be part, leading to an additional waiting time of up to 12 h. When calling at European container ports, shipping lines already reserve their place in a convoy and as such want to ensure that the vessel will make it in time to the Canal's entrance.

Chance

This includes unexpected waiting times due to weather circumstances, on route mechanical problems or unexpected waiting times at a bunkering site or port. For instance, it takes days, sometimes even weeks, before terminals in Europe and the US east coast recover from major schedule disruptions due to heavy weather on the Atlantic Ocean. Also on other trade routes weather conditions can seriously disrupt schedules and port operations.

As a liner shipping service consists of a set of roundtrips of individual vessels, delays in one of the segments of the roundtrip have cascade effects in the round voyage considered. The time buffers shipping lines build in their liner service schedules are typically very low. As a result, unexpected vessel waiting times in one port cascade throughout the whole loop. These intra-roundtrip effects can for example take the form of cascading late arrivals of a specific vessel at next ports of call. On top of this, delays for one vessel in the loop can also have an impact on round voyages of other vessels operating in the same loop. These inter-roundtrip effects might materialise by irregular service frequencies at a port in the loop (eg instead of one vessel call per week, an elevated variation in time intervals, for example 4 days, then 8 days, then 6 days, etc).

General notions on reliability and vulnerability

The container shipping business has become very competitive, and profit margins have become very small. In such an environment, the pressure is especially high to cut costs. Port fees and tariffs form only part of a carrier's total costs. Still, the tendency may be to put more emphasis on port fees and tariffs than on costs related to service time and reliability, because it is easier to measure the impact of port fees and tariffs on the carrier's bottom line than the impact of service time and reliability.

The reliability of a liner service network can be defined as the probability that one or more of its links does not fail to function, according to a set standard of operating variables. A non-functioning, or at best, badly functioning link will impose costs on the shipping line in terms of loss of time, additional operating costs or other costs as a result of delays and diversions. As mentioned earlier, the shipping lines' clients might also experience a loss of value, potentially resulting in claims filed against shipping lines depending on contract terms.

Reliability affects the degree of stability of the quality of service that a system offers. Taylor and D'Este (2003) note that vulnerability and reliability are two related concepts, but emphasise that network vulnerability relates to network weaknesses and the economic and social consequences of network failure, not so much the probability of failure. In the context of shipping services, vulnerability could be defined as the inability to supply adequate serviceability. As such, reliability focuses on the possibility of maintaining a link; vulnerability focuses on the possibility of disrupting or degrading a link. With reliability and vulnerability being two different concepts, a measure of reliability does not translate into a measure of vulnerability.

Schedule reliability and transit time reliability

A schedule is a published timing of a round-voyage of a specific ship. A delay does not necessarily imply low schedule reliability. A carrier can be confronted with delays of a vessel within a specific loop, but these delays will only lead to schedule unreliability from the perspective of the customers in case the carrier does not succeed in remedying the delays by taking appropriate measures (eg by increasing vessel speed on the main haul).

Similarly, transit time reliability is not the same as schedule reliability. Sometimes, a shipping line succeeds in guaranteeing the delivery of a container in the indicated port in time, while schedule reliability of the vessel is poor. A typical example is the unplanned shifting of cargo from one vessel to another vessel in view of getting the container in time at the port of destination. In other words, a shipping line wanting to meet the scheduled transit time from, for example, Shanghai to Rotterdam, but confronted with disruptions in the normal loop timings, might shift containers between vessels or opt for land feedering out of other ports.

Shipping lines have diverging strategies when it comes to dealing with schedule reliability and transit time reliability. One of the key decision variables in liner service design relates to buffers. The probability of not meeting reliability targets increases as time buffers in the system decrease. For instance, carrier MSC keeps time buffers relatively low. MSC does not have the best record when it comes to schedule reliability, but still manages to have a reasonable transit time reliability by managing loops and vessels in a creative way. The observed 'creativity' comes in the form of ad hoc changes to the order of port calls, the ad hoc transhipment of containers at relay ports in the Mediterranean and the seemingly random skipping of one or more ports of call during a round voyage. Alternatively, shipping lines like Maersk Sealand and ACL are very strict in respecting the scheduled times and the order of ports of call. Time buffers are sufficiently high to cope with unexpected disruptions. As a result, Maersk Sealand is reputed for its high schedule reliability and consequently also its high transit time reliability. Guaranteeing a high schedule reliability and a high transit time reliability has its price: the rates of Maersk Sealand are known to be substantially higher than low-cost carrier MSC.

The relative importance of each of the sources of schedule unreliability in the liner service schedules on the East Asia–Europe route is depicted in Figure 4. The shares are the result of a survey among shipping lines and alliances that operate one or more of the loops as listed in Table 1. The data relate to an observation period of around three months, that is, from early October 2004 till late December 2004. This period includes the traditional peak season leading up to the Christmas shopping period. A number of shipping lines in the survey underlined that schedule reliability during the period of observation was the lowest in years due to a combined effect of historically high export flows out of China, port congestion and vessel capacity shortages. As such, the delay conditions in this period were exceptional for the major container ports and shipping lines. While exact figures are not available, respondent shipping lines underlined that between 70% and 80% of the individual vessel loops where facing late arrivals in at least one of the ports of calls along the route. In other words, only 20% to 30% of the cases showed sufficient schedule reliability. Leading terminal operating companies confirm these figures. For instance, in the first quarter of 2005, reliability of ocean carriers at the Antwerp facilities of PSA Corp. was just 30%. Ports need a schedule reliability of more than 90% for efficient terminal planning. Often, stevedoring companies receive only 24 h notice of vessel delays.

Figure 4.

Sources of schedule unreliability on the East Asia–Europe route for the fourth quarter of 2004 (based on survey data for the loops listed in Table 1).
Source: Survey data

Full figure and legend (33K)

The logistics implications of poor schedule reliability can be considerable (eg late deliveries of time-sensitive goods). That is why freight forwarders and shippers factor reliability risk into their decision for a certain service. Reliability impacts on pickup and delivery costs and is extremely critical when working with tight time windows. But when it comes to schedule reliability in the eyes of the forwarder or shipper, things are not always as they seem. For example, a trucking company seeking to maximise revenue might utilising opportunity to charge their clients demurrage, for example, a charge for being delayed for alleged congestion at the port, although port access is unconstrained. This might affect the shipper's perception of schedule reliability and port reliability.

Dealing with delays

Shipping lines have a range of possibilities to lower the risk of low schedule and transit time reliabilities.

First of all, reshuffling the order of ports of call is common practice. In some cases, this coincides with discharging more import cargo at the first port of call combined with the transfer of containers over land to destinations near ports that will be called at much later than initially planned. For instance, when a shipping line decides to change the scheduled port order Le Havre/Antwerp/Hamburg/Rotterdam to Antwerp/Hamburg/Le Havre/Rotterdam then the bulk of the import cargo destined for French customers will exceptionally be discharged in Antwerp instead of in Le Havre.

Secondly, a shipping line might cancel one or more port calls to cut total port time and get the vessel back on schedule. Skipping a port of call can have considerable implications on the pattern and costs of land transportation. In case a specific container has a bill of lading for port A, but the cargo ends up in port B because the call at port A was exceptionally cancelled, then the shipping line has to arrange and pay for inter-port transport from B to A. Customer satisfaction might decrease in case cancelling port calls becomes more of a rule instead of an exception. The frequent skipping of specific port calls often preludes a schedule redesign.

A special case is the 'cut and run' principle, which could be an option when calling at tide-dependent container terminals. 'Cut and run' implies that the crane operations on a vessel are abruptly stopped so that the vessel can leave the port, even though there are still containers left on the stack that normally should be loaded onto the vessel. These containers will either wait for the next vessel or are transferred over land to a neighbouring load centre. The incentive for the application of the 'cut and run' principle typically lies in avoiding unproductive port time caused by low tide situations. In Antwerp, Maersk Sealand vessels from time to time cut operations and run to make sure they can benefit from favourable tidal windows. In case all loading/discharging operations would only be ready by the time the draft conditions on the river are less favourable, the vessel would have to sit and wait for the next high tide.

Thirdly, shipping lines might deploy other vessels to take over (in combination with delivery to hub). Delays of vessels might be compensated by the phase-in of other vessels into the published schedule. The delayed vessels are phased out to lay-by periods and are deployed again on demand. This supply on demand culture is causing periods where the vessel is out of service for an uncertain period.

Fourthly, shipping lines might speed up turnaround time at next port(s) of call in the loop to catch up and resume the schedule. A number of container terminals are known for their ability to make up time lost in other ports. In Europe, for instance, many shipping lines consider Antwerp as a kind of safety valve, where very high terminal productivities can be achieved in view of bringing a vessel back on schedule.

Finally, carriers might make up time by increasing vessel speed on the intercontinental trunk route. Bendall and Stent (1999) have referred to speed as a way to reduce roundtrip time. Increasing vessel speed incurs higher bunker costs. The use of fast vessels is associated with higher capital costs than for slower vessels, but once a vessel is in service, it is generally assumed that capital costs are not affected. The additional costs need to be counterbalanced by savings in time costs.

Vessel speed, bunker costs and schedule reliability

There is an obvious trend in the modern container ship designs towards higher speeds and increasing speed margin, primarily for maintaining a tight sailing schedule with good frequency and reliability. Baird (2001) identified two ways to estimate fuel costs: (a) the specific fuel oil consumption (SFOC) in grams per hp hour (g/bhph) multiplied by the normal utilisation of power to achieve desired service speed and (b) the stated daily fuel consumption in tonnes per day at desired service speed.

Fuel costs typically represent half of the total operating costs of modern container vessels (see Table 6). A container vessel and its engine are designed to run at the desired service speed at a reasonable level of daily fuel consumption. Increasing speed means that the vessel will reach the steep (exponential) segment of the fuel consumption/speed curve, meaning a disproportionate increase in fuel consumption. Lowering vessel speed brings savings but from a more marginal perspective. In order for a vessel to have 'healthy' fuel costs, ship operators stay within a speed range (see Figure 5). The form of the graph not only contains the complex interaction of the propeller and hull, but also the implicit fuel price. At speeds where the curve is relatively flat, operating speed can be increased with very little penalty. At speeds where the curve is steep, there are great benefits to be gained from slowing down. As a result of port congestion at the US West Coast, a number of carriers slowed down vessel speed on the trans-pacific route to 19 knots instead of the normal 22–23 knots. Preferred operating speeds are at those points on the curve where it starts to become appreciably steeper. However, in order to compensate for the cost per hour of sailing at a slower speed with the savings per hour in fuel costs, the shipping line needs to estimate its value of time and that of its customers (see the calculation of time costs of goods earlier in this paper).

Figure 5.

Example curve of daily fuel consumption/vessel speed.
Note: Data for CSCL Oceania, 8,468 TEU, 93,000 bhp.
Source: Based on data provided by Sea Span

Full figure and legend (43K)

Table 6 - Operating costs of panamax and mega-post-panamax ships.

Full table

CONCLUSIONS

Managing the time factor in the design and operation of liner services is an important challenge faced by shipping lines. Complex logistics networks have emerged demanding high frequencies, low transit times and high schedule reliability at the lowest possible cost. But at the same time, fast growth in cargo volumes in recent years has lead to port congestion becoming one of the main causes of disruptions in service schedules. Shipping lines are constantly balancing factors such as the risk of late arrivals and the minimisation of scheduled and actual transit times. They deploy a wide array of measures to ensure schedule reliability and transit times as much as possible.

This contribution examined the time factor in liner services design and operation. The paper analyses how shipping lines deal with the trade-offs linked to managing the factor time in liner service design and discusses the range of measures and planning tools container carriers deploy to maximise schedule reliability.

The analysis provided in this paper leads to some conclusions and challenges in view of further research. First of all, carriers' strategies in dealing with potential disruptions in service schedules differ substantially. This identifies a service variable that adds to market segmentation in liner shipping. Secondly, port congestion is the main source of schedule unreliability. During the Summer of 2004, shipping lines learned the high cost of port delays they had never expected. It has made shipping lines more susceptible to finding ways of anticipating possible delays, for instance by means of adding flexibility in their liner service networks. The risk of port congestion also partly explains why a large number of shipping lines are securing terminal capacity at key locations in their shipping networks, through minority shareholdings or joint ventures. Thirdly, this paper focused on intra-roundtrip effects of schedule unreliability on the loops between East Asia and northern Europe. In future research the scope of the analysis could be extended to include other trade routes, inter-roundtrip effects and network effects. As liner service networks have become more complex, the possible impact of service disruptions in one segment of the network on the network as a whole increases and could substantially lower the attractiveness of some types of liner service design.

References

1. Baird, AJ. 2001: A new economic evaluation of the hubport versus multiport strategy. Proceedings of the IAME 2001 Conference, Hong Kong. pp. 138–166.
2. Bendall, HB and Stent, AF. 1999: Longhaul feeder services in an era of changing technology: An Asia-Pacific perspective. Maritime Policy and Management 26: 145–159. | Article |
3. Cullinane, K, Khanna, M and Song, D-W. 1999: How big is beautiful: Economies of scale and the optimal size of containership. Proceedings of the IAME 1999 Conference, Halifax. pp. 108–140.
4. Drewry Shipping Consultants. 2001: Post-Panamax containerships – The next generation. Drewry Shipping Consultants: London.
5. Drewry Shipping Consultants. 2003: Annual review of global container terminal operators. Drewry Shipping Consultants: London.
6. Fagerholt, K. 2004: Designing optimal routes in a liner shipping problem. Maritime Policy and Management 31: 259–268. | Article |
7. Gilman, S. 1999: The size economies and network efficiency of large containerships. International Journal of Maritime Economics 1: 5–18.
8. Lago, A, Malchow, M and Kanafani, A. 2001: An analysis of carriers' schedules and the impact on port selection. Proceedings of the IAME 2001 Conference, Hong Kong. pp. 123–137.
9. Lim, S-M. 1998: Economies of scale in container shipping. Maritime Policy and Management 25: 361–373.
10. Meyrick, S. 2004: Structural changes bring new challenges. Port Strategy, January: 33–35.
11. Notteboom, T. 2004: A carrier's perspective on container network configuration at sea and on land. Journal of International Logistics and Trade 1: 65–87.
12. Robinson, R. 1998: Asian hub/feeder nets: The dynamics of restructuring. Maritime Policy and Management 25: 21–40.
13. Taylor, M and D'Este, G. 2003: Network vulnerability: An approach to reliability analysis at the level of national strategic transport networks. Proceedings of the 1st International Symposium on Transportation Network Reliability. Elsevier Science: Kidlington. pp. 23–44.
14. Verbeke, A and Notteboom, T. 2003: Studie naar de directe baten van de verruiming van de Westerschelde: een logistieke benadering (Study on the direct benefits of the deepening of the river Scheldt). Report commissioned by ProSes. Antwerp. December 2003.
15. Yap, WY, Lam, JSL and Notteboom, T. 2003: Developments in container port competition in East Asia. Proceedings of IAME 2003 Conference, Busan. pp. 715–735.