Exploring the effects of flexible inland network
selection for containers transported oversea: the concept
of synchronetwork
MSc Supply Chain Management & Technology Operations Management
University of Groningen, Faculty of Economics and Business
Bart Somhorst
b.p.somhorst@student.rug.nl
Student number: s3273393
Supervisor: dr. ir. S. Fazi, University of Groningen
Second supervisor: dr. N.D. van Foreest, University of Groningen
Word count: 13499
Date: 17-07-2019
AbstractThe increase of congestions at ports, and the restrictions on the use of truck as a mode for inland transportation, will require innovative future transportation systems. Literature on liner shipping focuses mostly on optimisations at the sea-side transport of containers. However, little research is conducted with regard to the integration between the sea-side transportation, and inland transportation. In this research, the concept of ‘synchronetwork’ is introduced, which suggests a system in which the inland transportation networks can be selected based on real-time market information. By doing so, more efficient and sustainable transport could be realized. The goal of this design science research was to develop a validated synchronetwork design. The findings indicate that a synchronetwork system could have potential, this is however under the specific assumption that in future scenario’s real-time market information for the inland transportation of containers will be available.
Acknowledgements
2
Table of content
1. Introduction ... 5
2. Literature background ... 7
2.1 Global supply chain of containers ... 7
2.2 Foreland container transportation ... 7
2.3 Inland container transportation ... 8
2.4 Container handling ... 11 2.5 Contribution to literature ... 12 3. Methodology ... 14 3.1 Research method ... 14 3.2 Problem analysis ... 15 3.3 Solution design ... 16 3.4 Validation ... 16
4. Problem analysis and solution design ... 17
4.1 Approach ... 17
4.2 Current operational process ... 17
4.3 Synchronetwork’s operational process ... 18
4.4 Factors influencing the synchronetwork concept ... 19
4.4.1 Vessel Stowage ... 19 4.4.2 Terminal handling ... 21 4.4.3 Administrative acts ... 23 4.4.4 Inland transportation ... 24 4.5 Necessary elements ... 26 4.6 Assumptions ... 27 5. Synchronetwork design ... 28 5.1 Functional architecture ... 28
5.1.1 A.0 - Core synchronetwork function ... 28
3
5.1.3 Function A.2 – Container handling function ... 33
5.1.4 Function A.3 – Hinterland function ... 34
5.2 Synchronetwork’s economic benefits ... 37
5.2.1 Scenario 1 – Small selection / little restowage ... 38
5.2.2 Scenario 2 – Small selection / lot of restowage ... 39
5.2.3 Scenario 3 – Large selection / little restowage ... 40
5.2.4 Scenario 4 – Large selection / lot of restowage ... 41
5.2.5 Distance threshold for cost savings ... 42
6. Validation ... 43
6.1 Design validation ... 43
6.2 Cost savings validation ... 44
7. Discussion & limitations ... 46
4
Table of figures
Figure 1 – Schematic figure of a container vessel (J. Christensen & Pacino, 2017) ... 11
Figure 2 – Schematic figure of a container overstowage (J. M. Christensen, 2017) ... 12
Figure 3 - Regulative cycle Wieringa (2009) ... 14
Figure 4 - Carrier haulage container transport ... 18
Figure 5 - Synchronetwork transport ... 18
Figure 6 – Hierarchy 'Core synchronetwork functions' ... 29
Figure 7 - Interactions 'Core synchronetwork functions' ... 30
Figure 8 - Hierarchy 'Synchronetwork transport function' ... 31
Figure 9 - Interactions 'Synchronetwork transport function' ... 32
Figure 10 - Interactions 'Discharging port selection function' ... 33
Figure 11 - Hierarchy 'Container handling function' ... 33
Figure 12 - Interactions 'Container handling function' ... 34
Figure 13 - Hierarchy 'Hinterland function' ... 35
Figure 14 - Interacions 'Hinterland function' ... 36
Figure 15 - Interactions 'Route information function' ... 36
Figure 16 - Scenario 1 ... 38
Figure 17 - Scenario 2 ... 39
Figure 18 - Scenario 3 ... 40
Figure 19 - Scenario 4 ... 41
Figure 20 - Cost savings distance threshold ... 42
Figure 21 - Hierarchy ‘Perform liner shipping in synchronetwork context’ ... 56
Figure 22 – Hierarchy 'Transport container (synchronetwork style)' ... 57
5
1. Introduction
In recent years, sea ports are competing more and more with each other based on their inland transport network (van der Horst & van der Lugt, 2011). This concerns not only neighbouring ports, but also distant ones, since an efficient inland transport network can easily reach any destination. For example, the port of Piraeus is a competitor of the port of Rotterdam, due to fast connection to Eastern Europe. Therefore, it is clear that an efficient inland transport system is getting a decisive factor for logistic shippers to choose a specific discharging port rather than another. Things like the lack of storage space, and unbalanced transport networks, call for a different approach with regard to container distribution. This while realizing more sustainable and efficient inland transport.
Ideally, the choice for discharging a container at a specific port could be driven by the availability of transport services, and the status of the network. For example, a shipper may decide to discharge its container to a different port than originally planned, because of the availability of more efficient and sustainable modes for inland transport, or because of unexpected congestion at the original port of destination. Another benefit that could arise would be that ports are relieved in case their networks are highly utilized or broken. This would ensure a better service to the shipper. However, the choice of flexible network selection has its own challenges and these concern mainly logistics and organizational costs. For example due to the unfavourable position of a container on a ship, re-stowage of certain containers could be needed before being able to discharge a specific container. Besides of that, the rescheduling of a container’s route may require a lot of coordination and additional effort from the shipping line or the decision maker in general. In addition, shippers need to have a real-time overview of current prices and network status.
To our knowledge the proposed concept, named “synchronetwork” in this thesis, is not applied since discharging ports typically remain a fixed decision. However, the literature proposes several ideas that fit into the concept. For example, Tran (2011) already addresses that there is a gap in literature with regard to liner shipping, and its integration towards the hinterland networks destination. Lee & Song (2017) also emphasize on the importance of this integration, besides of that they mention that it would be of great interest to design an intermodal network, where inland, as well as foreland transportation is taken into account. Due to the hinterland expansion of major ports, ports increasingly compete with each other for the same hinterland (van der Horst & van der Lugt, 2011). And because of the infrastructure utilization advantages a synchronetwork design promises to realize, the overall competitiveness of ports within a certain region will increase. These infrastructure advantages are realized by influencing inland transportation network selection, based on real-time inland network information.
6 flow of containers and decisions within a synchronetwork transportation system. This leads us to the following research question:
“How should a validated functional model for synchronetwork container transportation be shaped?”
The research is conducted by the use of Design Science Research (DSR). DSR is chosen as a method, since it is able to design and validate models, while also contributing to literature (Van Aken, Chandrasekaran, & Halman, 2016). In doing so, first the factors which are influencing such a concept are determined. This is done by interviewing experts on the field of liner shipping. Based on their expert view with regard to factors which will have an influence on such a design, we were able to determine the necessary elements for a synchronetwork transportation system. These elements are used in order to develop the synchronetwork design. The designed concept consists of multiple hierarchical layers, where the interactions within each layer are discussed in detail.
7
2. Literature background
This chapter consists of the literature review with regard to concepts in the literature which have brought us to our literature gap. Section 2.1 focuses on the global supply chain of a container, where the sea transport, as well as inland transportation of a container are mentioned. Section 2.2 gives further explanation on the foreland transportation of a container, which refers to the sea transport stage. Once a container is transported oversea, it is entering the hinterland, which is the focus of section 2.3. Section 2.4 focusses on the handling operations with regard to the transportation of a container. And finally section 2.5 gives the contribution of the paper to current literature, and what gap is addressed.
2.1 Global supply chain of containers
The supply chain of a container can be divided in two transportation stages. The first stage is the foreland transport, where the foreland is referred to as the overseas area with which a port carries out trade. This foreland transport is typically carried out via liner shipping companies. Liner shipping is involved in a whole network of ports, where each liner vessel has its own route in which it passes several ports (Tran & Haasis, 2015). After the foreland transportation, a container ship will enter a port, where the transported containers will be made ready for inland transportation. The hinterland of a port is referred to as the transport network over which cargo is imported/exported into or from a seaport (Talley & Ng, 2017). With concern to inland transport, Iannone (2012) states that there can be made a distinction between merchant haulage, where the inland transport is regulated by the consignee of the container, and carrier haulage, where the inland transportation, just like the foreland transportation, is in charge of the shipping line. Carrier haulage is better able to deal with flexibility on the spot, compared to merchant haulage, but is also more costly. The reason why carrier haulage is better in dealing with flexibility, is because decisions can be made centrally, which allows for a certain degree of flexibility in decision making. In this thesis, we will refer to the case of transporting containers under carrier haulage. Since this will give us the ability to make centralized and autonomous decisions with regard to a container’s transportation route.
2.2 Foreland container transportation
8 aspects are both directly related to the transportation costs. However, the effect of oceanic distances has a great impact only in case ports are located at opposite coasts. When ports are located near each other along one coast line (for example in the Le-Havre/Hamburg line), the effect of oceanic distance is minimized, due to the fact that oceanic distances are regarded as near-equivalent in such case. As mentioned by Wiegmans, Hoest, & Notteboom (2008) port selection is based on strategic reasons, while the terminal selection is based upon the financial reasons. In port selection there is mostly decided upon the availability of connections, reasonable tariffs, and immediacy of consumers (large hinterland). Terminal selection is based upon the handling speed, handling costs, reliability, and hinterland connections. Mulder & Dekker (2014) combined the fleet-design, ship-scheduling, and cargo-routing problem. By doing so, total revenue is maximized. Tran (2011) studied port selection in liner shipping. The model determined the port’s visited in a route, at which order, and where to load/unload each shipment. This while minimizing total costs, based upon the ship costs, port tariffs, inland transport cost, and inventory cost. This model was novel to literature since it took the inland transportation costs as a variable of importance in determining port selection. Liu, Meng, Wang, & Sun (2014) presented a model where the fore- and inland transportation are integrated. They propose to design seaborne liner services in an intermodal way, rather than designing it purely based on port-to-port demand. This means that the inland destinations of containers are taken into account as well, in determining the routing and port selection of the liner vessel.
2.3 Inland container transportation
The inland transportation of containers from and to a seaport is a complex process, which has a great effect on the internal costs of transportation. For inland transportation of containers there are used three general types of modes, namely road, rail, and barge. With regard to the inland transportation, there is made a distinction between the fundamental and competitive hinterland. Where the fundamental hinterland is geographically located in such a way, that it is always served by a specific port. The competitive hinterland is described as the market areas over which several ports compete intensively with each other for business (Rodrique, Slack, & Notteboom, 2017). For the synchronetwork concept, the competitive hinterland will be of interest, since decision making will be based on the inland network selection of different ports.
9 2035 with regard to their modal split, where in 2014, 36% of their containers were transported by barge, 11% by train, and 53% by truck. In 2035 this should be at least 45% by barge, at least 20% by train, and at most 35% by truck (Fazi, Fransoo, & Woensel, 2015).
Increasing the share of train and barge transport is important for ports, since it effects a port’s competitive position (van den Berg, 2015). Train and barge transportation is however less flexible in compare to truck transportation. Where train transportation is even less flexible than barge transport, since it shares its infrastructure with passenger trains (Behdani, Fan, Wiegmans, & Zuidwijk, 2014). However, the capacity of the infrastructure for barge and train transportation is for most ports more than sufficient, while it is not utilised to its full capacity (van den Berg, 2015). This is mostly due to the complexity with regard to both transportation modes. Trains are mostly pre-scheduled, and dealing with long distance inland destinations, which makes them static and not flexible in planning. Besides, barges and trains deal with complex scheduling because they consolidate containers which differ with regard to their delivery window (Fazi et al., 2015). This higher complexity relates to an increased chance of coordination problems with regard to the (un)loading of containers, and therefore lower capacity utilisation of both modes (van der Horst & van der Lugt, 2011). In order to decrease the drawbacks of more sustainable inland transportation, and realize economies of scale, the alternative inland transportation modes need to be promoted (Fransoo & Lee, 2013).
From a tactical point, better utilisation of more sustainable transportation modes should be realized by concepts such as intermodality and synchromodality. These concepts realize more flexibility and better utilization of different modes for inland transportation (van Riessen, Negenborn, & Dekker, 2014). In case of synchromodal transportation, the shippers book their transport service ‘mode-free’, which means that they do not need to specify their inland transport mode in advance. This will result in more efficient and flexible transport, because inland transport companies will be able to bundle flows of goods which originate from different customers (Pfoser et al., 2016). This is to some degree similar to the synchronetwork concept, where shippers have the freedom to flexibly choose the inland transportation networks through which they transport their containers. This decision will be made close to arrival, based on real-time information.
10 transportation could have a significant impact on the total transportation costs. Notteboom (2004) and Roso (2015) state that liner ships are increasing in size, which makes the foreland transportation cheaper. As a result, the potential for substantial cost improvements now shifts towards the inland transport part. An integrated system would be needed in order to reach the full potential of such improvements. In order to realize a more integrated system, the switch needs to be made from a push to a pull logistics system. Where integration between suppliers, manufacturers and distributors will be needed. In case of realizing such levels of integration, a system will be able to achieve more efficient transport (Rodrigue & Notteboom, 2010).
In case of a synchronetwork, it would be essential that a container is in carrier haulage. This because in that case the shipping line will have full control over the container transportation, and may decide at its discretion to transport a container via a different network. Most of the containerized cargo is however transported under merchant haulage. In case of merchant haulage, synchronetwork decision making would be way more complex, and practically infeasible. This because the inland transportation network selection would then need to be discussed with the forwarding party. This will lead to way more complexity, and besides of that, there will be no economic advantages for the shipper himself. There is estimated that nowadays less than 30% of the containers related to the market share in Europe are transported under carrier haulage (Legros, Yann, & Fransoo, 2019). Shipping lines try to increase this percentage, since they prefer to have full control over their container flows, as is realized in case containers are transported under carrier haulage. The percentage of carrier haulage is also shipping line specific, where for example MSC is more focussed on port-to-port transportation, and Maersk has a stronger focus on the door-to-door transportation (van den Berg, 2015). As a result, only part of a ship’s freight qualifies for a synchronetwork transportation approach.
11 for these modes will even grow more. Since containers can now also be assigned to such a mode based on real-time market information. In making the trade-off for synchronetwork decision making, the selected containers original mode of transportation should be less cost efficient than the offered alternative. Where the transportation cost savings will be determined based on the transportation mode and inland destination distance (van Riessen et al., 2014).
2.4 Container handling
Christensen (2017) mentions three different stages of container handling with regard to the transport of a container. These stages are the vessel stowage, terminal handling, and inland transportation. In figure 1 a container vessel is shown, where containers are stacked in bays as can be seen. Each container is assigned to a specific slot within a container stack. The on-deck and below-deck containers are separated by a hatch-cover. This hatch-cover functions as a leak-proof cover so no fluids will be able to get below deck (J. Christensen & Pacino, 2017).
The first stage of a container’s transport deals with the loading of a container on a ship. In order to realize fast and efficient transhipment of containers, good distribution of the containers over the ship is needed (Vis & de Koster, 2003). Therefore, from an operational level, stowage plans are made. The stowage planning deals with the problem related to the allocation of containers on a ship (Mulder & Dekker, 2016). In developing these stowage plans a lot of constraints need to be taken into account, which makes it a very complex planning process. Also things like ship stability on sea, and stability during vessel discharging need to be taken into account. The assignment of containers on a specific slot is referred to as the cargo mix problem. J. Christensen, Erera, & Pacino (2019) developed a matheuristic approach to solve the cargo mix problem, which is able to take what-if scenarios into account due to a stochastic setting. Cohen, Coelho, Dahan, & Kaspi (2017) developed a matheuristic software system to create vessel stowage plans. This software takes the upper as well as the lower part of a container vessel into account. Where the containers are first devoted to sections in bulk, after which the specific container slots are assigned to each container. There is made a distinction between different types of containers, which are reefers (refrigerated), empty, dangerous, standard, and out of standard (Junqueira, 2011). Parreño, Pacino, & Alvarez-valdes (2016) developed an algorithm which introduced the inclusion of separation rules for dangerous cargo. In our research the focus will be on the standardized containers.
12 Once a container ship arrives at a port and is assigned to a berth place, where a terminal can start discharging the containers. The unloading is conducted by quay cranes, which lift the containers of the ship and puts them on internal vehicles (vehicles belonging to the terminal). It happens that in case a container is positioned below container(s) which need to be unloaded later, the container(s) on top need to be restowed first (Christensen, 2017). In case such restowage handlings are needed, there is dealt with overstowage. Ding & Chou (2015) developed a heuristic algorithm which should reduce the total amount of shifts in creating stowage plans. Where shifts refer to the restowage handlings which are needed in order to reach targeted containers. Literature distinguishes two types of overstowage, namely stack and hatch overstowage. These two types of overstowage are depicted in figure 2 below:
Figure 2 – Schematic figure of a container overstowage (Christensen, 2017)
In case of stack overstowage, the targeted container is located on-deck, and only containers located on top of the targeted container need to be restowed. Hatch overstowage refers to the situation where the targeted container is located below-deck. This means that the containers on-deck, so above the hatch cover, need to be restowed first, in order to be able to remove the cover and discharge the targeted container. Once a container is discharged, it is stored in the yard where it can be picked up for inland transportation (Junqueira, 2011).
2.5 Contribution to literature
14 Figure 3 - Regulative cycle Wieringa (2009)
3. Methodology
3.1 Research method
Design science research (DSR) consists of two activities, first of all it is used to build certain designs, artifacts, and processes, where it is also used for the validation of these constructs (Hevner, 2014). DSR consists of two important components, where one deals with the practical problem which is experienced, and the other focuses on the intervention or solution to the problem (Van Aken, 2007). In order to be able to design a solution, a deep understanding of the problem situation, as well as the systems current way of working is necessary. After the necessary understanding of the concept is gathered, a solution to the problem can be designed (Van Aken et al., 2016). The developed solution should be able to answer the following research question:
“How should a validated functional model for synchronetwork container transportation be shaped?”
Wieringa (2009) provides guidance on the process of designing a solution, in advance of the implementation phase. Four steps are of importance, which forms a regulative cycle (Figure 3). In our research, we focussed on the first three steps in advance of implementation. Based on these three steps of the regulative cycle, four more sub-questions have been determined:
1. Identifying the problem;
- What is known from literature with regard to flexible port selection in container transport? And how would such decision making look like?
- What are the necessary elements of a synchronetwork concept? 2. Design phase;
- What is a possible design, taking all necessary elements into account? 3. Validation of the design;
15
3.2 Problem analysis
The investigation of the problem was partly addressed in the second chapter. Where the gap in literature was described, which was used as a base for the concept proposed in this thesis.
In order to investigate what factors have an influence on the synchronetwork concept, semi-structured interviews with experts on the field of liner shipping have been conducted. Semi-structured interviews have been chosen as a research method, since these can give a certain degree of flexibility during the interview. This extra flexibility could result in more insight into an aspect(Voss, Tsikriktsis, & Frohlich, 2002). In order to analyse the data gathered by the interviews, it was first transcribed. The data has been coded in order to analyse it in a structural and systematic way. Two types of coding were used, namely deductive and inductive coding. In case of deductive coding, the codes are derived from a pre-set framework, where with inductive coding the coding themes emerge during the discussions (Fereday & Muir-Cochrane, 2006). Christensen (2017) mentions three stages in container transport, namely the vessel stowage, terminal handling, and inland transportation stage. These stages form the base for the deductive coding. Since semi-structured interviews have been used as a method, for which there was a certain degree of ‘by-catch’ information. This information has been coded in an inductive way, by first categorizing the information into different categories. This procedure is called open coding and results in 1st order concepts. By searching for relationships between these 1st order concepts, 2nd order codes
were derived. This procedure is called axial coding. For the 2nd order codes, overarching concepts were
determined, which is called selective coding (Strauss, & Corbin, 2006).
People with several backgrounds related to liner shipping were interviewed. Based on their information, the processes within a synchronetwork system have been defined, as well as the factors influencing these processes. The liner shipping experts had the following background (Table 1):
Table 1 – Interviewees background
Function Background
Project director and former COO Liner shipping
Equipment flow operator Liner shipping
PhD candidate on container supply chain Literature expert
Network manager Container transport
16
3.3 Solution design
By combining the information gathered in the different interviews, we were able to develop a list of necessary elements. These elements need to be taken into account in designing the synchronetwork concept. The design phase refers to the ‘solution design step’ of the regulative cycle as described by Wieringa (2009).
The design is made by the use of a program named Vitech Core. In here a functional architecture of all processes is created. Where a process is assigned to a function, and certain processes are again decomposed by smaller sub-processes. In order to visualize the functional architecture, an IDEF format is used. IDEF stands for Integrated Definition, and shows the interactions between the different functions.
3.4 Validation
17
4. Problem analysis and solution design
4.1 Approach
In the first two chapters we have introduced the synchronetwork concept, where inland transportation networks can be flexibly selected based on real-time market information. Since the synchronetwork concept is new to literature, the regulative cycle of Wieringa (2009) is used in order to provide guidance on the concept’s designing process. The first step of the Wieringa (2009) regulative cycle deals with the problem analysis. The questions related to this first step are:
- What is known from literature with regard to flexible port selection in container transport?
And how would such decision making look like?
The need for a design, based on the literature gap, is partly addressed in the first two chapters. To further analyze the problem, it is of great importance to have a good understanding of the related processes and the problem situation itself. In order to obtain this understanding, literature with regard to container handling processes is reviewed. Besides, the current and synchronetwork way of transporting a container is determined by means of interviews. Finally, the factors which are affecting synchronetwork decision making are determined. These factors are used in order to determine the design’s necessary elements. The necessary elements are considered of great importance in designing the synchronetwork concept. Also the design’s assumptions are determined and given in the last paragraph.
4.2 Current operational process
Based on the stages as described by Christensen (2017), the current operational process of container transportation can be sketched. Experts also validated the sketched process. First of all, containers can be booked under several agreements. In case of carrier haulage, the foreland as well as inland transport is arranged by the shipping liner (Fereday & Muir-Cochrane, 2006). Once a container is booked, it is loaded at its port of origin, from where it is stowed on a ship. The stowage happens according to a stowage plan, where besides a container’s attributes with regard to its inland logistics, also things like the container weight and size are taken into consideration. With regard to the booking moment of inland transportation of containers under carrier haulage, experts said the following:
“The moment a container gets loaded, everything with regard to its routing is booked.”
18 the ports hinterland, the container is transported via its pre-booked mode to its inland place of destination. The lay-out of the process is given in Figure 4:
Figure 4 - Carrier haulage container transport
4.3 Synchronetwork’s operational process
During the expert interviews the decision flows with regard to transporting containers under a synchronetwork design were discussed. There was determined that the decision process was based on three decisive factors, namely the total costs (including transportation and handling costs) and the delivery window of the container. The concept as
discussed with the experts is described below.
In the proposed concept of a synchronetwork, there would be a decision point with regard to whether it would be a better alternative to transport a container via a different inland transportation network. Since containers are currently distributed based on their original given port of destination, they do not take things like real-time market information, congestion or network utilization of the different inland networks into account. The incentive for taking such a decision should be cost driven in order to make it interesting for the liner shipper. Therefore, the decision to switch from inland transportation network is made in case the savings on inland transportation outweigh the extra required handling cost. These savings are based upon switching to a cheaper, more sustainable mode of transport, while complying with the container’s delivery window. The schematic process of distributing a container in a synchronetwork system is depicted in figure 5. The next section will focus on all factors influencing a synchronetwork system.
19
4.4 Factors influencing the synchronetwork concept
Now we have a good understanding of literature, and determined how the synchronetwork decision making differs from current system’s decision making. It is time to focus on the factors influencing the synchronetwork decision making. By doing so, the following question can be addressed:
- What are the necessary elements of a synchronetwork concept?
Deductive coding is used in order to categorize the information gathered during interviews with experts on the field of liner shipping. For the deductive coding process, the three operational stages as addressed by Christensen (2017) of container transport are addressed. These stages are the vessel stowage, terminal handling, and inland transportation. Where vessel stowage refers to the way in which containers are stacked in a ship. Terminal handling deals with all container movements necessary for containers to get discharged. And finally the inland transportation deals with the pricing and inland transport of a container. 2nd order codes have been created which are categorized based on these three sections. Our
interview protocol (Appendix A) took the three stages of Christensen (2017) as a base, where for each stage there was asked for the factors which would influence the synchronetwork system. Through the inductive coding of the ‘by-catch’ information, a fourth section was added. This fourth section is with regard to the administrative acts which are required when making a synchronetwork decision. Below, the factors which will influence a synchronetwork system are discussed. For each factor, quotes related to what was stated by experts are mentioned.
4.4.1 Vessel Stowage
Before a ship leaves from its origin, all containers need to be stowed onto the ship. The way a ship is stacked has great impact on the handling in a later stage of the container transport. Because of that, factors influencing the vessel stowage in case of a synchronetwork system, are regarded as of importance.
The first factor of importance with regard to vessel stowage deals with the location of a container on a liner ship. Experts mentioned that reasonable handling costs are made in order to restow or discharge a container.
“Only containers in top stow would be of interest to look at.”
Top stow containers will require relatively little restowage in order to get discharged/re-allocated. So for containers to qualify for the synchronetwork concept, they should have a top stow position.
20 otherwise not have been touched, there will be a negative impact on portstay, since extra handling will be needed.
“The discharging terminal could also be having difficulties when picking a single container in bays that otherwise would not be touched, again leading to an increase in portstay, and negative impact on the shipping line.”
Besides of the container location on a ship, the stacking method is also a factor of great importance. The stacking method refers to the way in which containers are stacked on a ship. Certain stacking methods are applied when loading liner ships, and in case of a synchronetwork system, a stacking method which fits the synchronetwork requirements needs to be used.
“The only way to achieve this is to ensure that containers are stowed in a vertical stack.”
“You would have to stack with the same container attributes rather than discharge port, as you want the latter to be flexible.”
In case a container is discharged in a port located further in a route, it still needs to be in top stow position to get discharged. Potentially the best way to achieve this is by stacking containers in a very specific way, namely by the use of vertical stacking. The vertical stacking principle works as follows: in case a container enters the system, it will be stowed onto a pile from which the first-leaving-container is regarded as of the ‘same type’ as the container which is entering the system. If no such a pile exists, the policy will assign the entering container to an empty pile. In case no empty pile exists in the system, the container will randomly be assigned to a pile (Gharehgozli, 2012).
Another comment made by an expert deals with the creation of stowage plans. When loading the containers at the port of origin, the difference could be made between containers which are regarded as ‘high potential’ for synchronetwork distribution, and containers which are most certainly fixed to the pre-determined port. When keeping this into account when stowing the ship, the potential benefits of the system may rise.
“It could be interesting to account for containers which will probably qualify for synchronetwork transport, in making the stowage plans for the ship.”
21
“And even if you stack vertically, there would be a number of extra restows that someone would have to pay for, just to gain access to the twistlocks.”
Alternatives should be thought of to current twistlocks, because the extra restows which are needed in order to get access to the twistlocks of certain containers will make the concept practically infeasible from a cost perspective.
4.4.2 Terminal handling
Once a container ship enters a port, it will get discharged by a terminal. These terminals have a great impact on the handling costs which need to be made in order to change a container’s port of discharge. In case a container is not located on top of a pile, the containers which are ‘blocking’ the targeted container need to be restowed first. The restowage of containers is done by lifting the ‘blocking’ containers and replace them to a spot where they are (temporarily) stowed. This process will require certain operational cost with regard to the replacement of the container with the quay crane asset.
“Just imagine that you have your 10 containers stacked in between containers for the first port and you are planning to discharge in the last port of the rotation.”
“A shifting or restow is typically is certainly in Europe quite costly. We are talking 1 container, 2 moves (off/on), and a move cost in excess of 100 per move.”
These costs should be taken into consideration in making the trade-off between inland transportation cost savings, and extra handling costs.
The restowage of a container will have significant costs, where the shift of one container may even result in 200 euro of handling costs. The total restowage cost can therefore result in very high handling costs. If 5 containers need to be restowed in order to discharge a container at a prior port in a route, there is already 1000 euro of handling cost that are made. This will easily outweigh the inland transportation cost savings.
Another factor deals with the fact that the terminal will need to have the specific stow codes of specific containers. For example in case restows are required, or when a container needs to be discharged at a different port. Expert views on this point variated a bit, some said:
“The terminal would have to have the special stow codes for the containers, and they do not like that.”
While others stated the following:
22 However, there was agreed that in case the handling of containers would be fully automated, the impact of needing to know the specific stow codes would be diminished. The biggest concerns were however with regard to the loading sequence. Restowing containers will complicate the loading sequence, while also taking up space. It would therefore be of great importance to manage these types of actions very well.
Comments were also made with regard to what type of terminal is handling the goods. In case the terminal is dedicated to a specific liner shipper, the costs that will need to be made for restowing a container are purely operational. Handling costs when re-allocating a container will therefore have less impact on liner shippers with dedicated terminals.
“You could make a difference between multi-user terminals, and dedicated terminals. Where in case of dedicated terminals, the terminal is seen as a tool by the shipping liner in order to transport containers as efficient and profitable as possible.”
The container weight and size were mentioned as having a big influence on the handling of containers. Since containers have different weights and sizes, they cannot just be stacked on each other. In making the stowage plans things like weight and size need to be taken into account in order to generate safe and feasible stowage.
“And that is assuming that we are talking about cargo of the same weight and container size, if that is not the case it complicates things even more.”
Stacking rules exist, which prescribe that 40 feet containers need to be stowed on top of 20 feet containers. If done the other way around, the construction of the 40 feet container may collapse (Christensen, 2017). Another example is that heavy containers need to be stowed on the bottom of a stack, since the construction may get instable and collapse in case of them being stacked on top. Such kind of stacking rules complicate the re-allocation of different weight and size containers.
Vessel stay was also mentioned as a factor which had indirect effect on transportation costs. In case of repositioning a container, extra time by the shipper is needed for discharging their vessel. In case this fits in the time-schedule of the ship, there will be charged no extra costs. However, in case extra time is needed, significant extra costs should be taken into account. And besides of that, in case of not meeting channel passages and fixed windows of a next port, even higher costs may occur.
23
4.4.3 Administrative acts
Now we will discuss the administrative acts which need to be in place when having synchronetwork decision making.
First of all the stowage instructions and load plans need to be sent 18 to 24 hours before vessel arrival to the terminal. Changes to these plans are basically not allowed, which means that before sending these, the decision whether to change a container’s port of destination needs to be made. Beside of that, there should also be enough time left in order to adapt the stowage instructions and load plans based on these changes. Because of that, a decision with regard to the inland transportation network has to be made 2-3 days before vessel arrival.
“We receive workplans (stowage plans) 18-24 hours before vessel arrival, and changes to these are basically not allowed after such time.”
“Decisions need to be taken 2-3 days before vessel arrival, as stowage instructions and load plans need to be prepared and sent to the terminal.”
Besides the stowage and load plans, documentation with regard to customs is also required. Customs require specific documentation with regard to the containers which are entering a port system. This documentation is country specific, and needs therefore to be adapted in case a container is discharge at a port located in a different country.
“Pre-registration of a container at the customs is always needed. So when signing up a container for transport via the Netherlands, but discharging it in Germany, extra administrative handlings will be needed. This might also result in administrative extra costs.”
Since the shipping liner arranges the whole transportation of a container when being transported under carrier haulage, there should also be taken a look at the port restrictions. However, according to experts they pose no restrictions.
“There would be no problem with bringing an extra container into the port system. This is a choice which can be made by the shipping liner itself, and forms therefore no restriction. This is however under the condition that there has been internal communication with regard to this decision.”
24
“A consignee gives the transport out of hand on purpose in case of carrier haulage. Because of that, it will be unlikely that the consignee will give restrictions with regard to the container’s port of discharge.”
However in case the container’s transportation route changes, it would be discreet to keep the consignee up-to-date with regard to the container’s routing.
4.4.4 Inland transportation
Finally, once a container has been discharged, it can be made ready for its transport towards its inland destination. The inland transport will be a decisive factor in whether or not to transport the container via a different inland network.
The inland transport pricing is a decisive factor in determining whether or not a container is being transported via a different inland transportation network. Due to the fact that larger liner shipping companies transport such a great amount of containers, they all have their inland transport arranged with contracts. Besides of that they use a system which calculates each route based on fixed prices, so in case of carrier haulage, the inland transport is automatically booked by this system. This also means that for larger shipping companies it is very unusual to take another look at the inland transportation cost, and see whether there is a cheaper alternative available on the market.
“Shipping volumes are of such a large quantity, that everything is already arranged by contracts. Because of that, we never re-asses the inland transportation of our containers. This makes that there does not exist something like a ‘spot market’.”
Although spot market prices aren’t available for large liner shippers, they apply to a certain degree to the ‘small players’ in the market. However, when looking at this transport itself, it is not 100% sure that the booked transport mode is actually going to transport that container. The bad service and non-transparency in the market makes that a spot market pricing for inland transportation does not exist.
“Such kind of spot market does not exist for big liner shippers, and due to non-transparency of the market it would be difficult to realize such a system in practice. Besides, we are dealing with large volumes. In case of shippers which deal with smaller volumes, the inland transportation could be arranged on a more spot-rate based way. This because they handle containers on an individual level.” “I believe that prices are a product of the competition in the market, and small players they for sure pay the list price, and on top they are the ones to receive the worst service.”
25
“For barge and train transportation there apply long-term contracts, where cost advantages are realized by transporting larger volumes. Not by competition between different providers.”
Barges and trains are very often not fully utilized due to the fact that they will not wait another day for an extra batch of containers. The better utilization of these modes is a field with a lot of potential, which is also addressed in concepts like synchromodality. By being able to make container distribution decisions on real-time information, these more efficient and more sustainable modalities can be better utilized. While the total inland transportation costs drop at the same time.
“Modalities like barges and trains are often not fully utilized, since they will not wait for the batches of containers which were also booked on them. At the same time, other containers which are now distributed via a different way could have been transported by that barge or train. Better information sharing will be needed in order to align containers with these transportation modes.”
Large liner shipping companies do not look at containers on an individual level, and because of that the system will currently be hard to realize in practice. Although smaller shippers will tend to look more at containers on an individual level, there do not exist programs yet which automate the price compare. Platforms will be needed where it will get easy to compare the inland transportation costs. TEU-booker is such a kind of platform used for merchant haulage. By comparing transportation options based on real-time information, decisions can be made with regard to cheaper and more efficient inland transportation options.
“It will be expensive to make a trade-off for every container at an individual level. However in case you would be able to (partly) automate it, and target container selections, it will take less effort and will be more efficient.”
26
4.5 Necessary elements
Based on the previous section, the necessary elements for a synchronetwork concept are determined. The following set of requirements can be derived from the previous paragraph:
- Container selection - Container stacking - Container handling - Administrative acts
- Real-time market information
27
4.6 Assumptions
Assumptions need to be made in designing the synchronetwork concept, this in order to account for certain scenarios and beliefs that are currently present in the real-world. Specific aspects of the design will not be in line with current practice, and therefore assumptions are set for these. The assumptions are listed below.
Assumption 1:
The first assumption made is with regard to the container stacking. The containers will need to be stacked according to the vertical stacking principle in order for the synchronetwork to be able to reposition containers. Therefore there is assumed that vertical stacking is used in the design.
Assumption 2:
Our design is assuming no disruptions in the system, since this will add a great amount of complexity. Therefore there is assumed that time-schedules are met, and no other unforeseen disruptions occur.
Assumption 3:
The flexible inland transportation network selection will be based on real-time market information, where the decision can be made to re-allocate a container to a more efficient transportation mode. In our system there is assumed that the market is transparent, so that containers can be assigned to a different inland transport mode based on its availability of free slots. This means that information with regard to network status and pricing will be available.
Assumption 4:
The decision making in a synchronetwork design should be made centralized. In order to do so, we assume that containers are transported under carrier haulage, and that liner shippers are fully in charge of decision making.
Assumption 5:
28
5. Synchronetwork design
This chapter will deal with the design phase of the Wieringa (2009) regulative cycle. The following question is related to this stage:
- What is a possible design, taking all necessary elements into account?
The design of the synchronetwork can be divided into two different stages. First a list with necessary elements is created. This list is mentioned in the previous chapter. The elements are considered as the critical factors which need to be in place for the synchronetwork concept. In this chapter, we will focus on the design’s functional architecture as well as calculations with regard to the design’s potential economic advantages. The functional architecture has been based on the information which was gathered and discovered during the problem analysis phase, and is shown in the section below. In the second section different cost saving calculations with regard to different synchronetwork decision making scenarios are given. Both the functional architecture, as well as the calculations will be validated in the next chapter.
5.1 Functional architecture
A functional architecture is defined by a set of functions which together perform the main function of the designed system. In our specific case, the main function is the transport of containers by the use of a synchronetwork system. The top hierarchical layers of the functional architecture can be found in Appendix B, where in the section below the functions with regard to synchronetwork decision making are discussed. This is the core synchronetwork function, and its sub-functions.
5.1.1 A.0 - Core synchronetwork function
29 Figure 6 – Hierarchy 'Core synchronetwork functions'
The interactions between the different sub-functions are depicted in Figure 7. Once container information with regard to its destination, delivery window, and stowage position is available, the core synchronetwork decision making can start. The synchronetwork transport function will by the use of inland routing possibilities, and specific container information, make a decision with whether to distribute a container via a different inland network. The different outputs deal with the container’s total transport costs, specific documentation, and certain route and delivery information. Stowage and load plans need to be adapted in case of distributing a container via a different transport network. The container handling function will change these plans based on the required handlings for each container. The discharge info it creates, is used by the liner shipper for discharging or re-allocating specific containers. The hinterland function determines on the one hand all possible inland routings for a container, while on the other hand collects routing information and details which will be needed for consignee information in case a container will follow a different route as pre-determined. In the next sections we will focus on the interactions between the sub-functions of A.1, A.2, and A.3.
A.0 Core synchronetwork... Function A.1 Synchronetwork transport function Function A.2 Container handling function Function A.3 Hinterland function Function hier Core synchronetwork functions
30 Figure 7 - Interactions 'Core synchronetwork functions'
Consignee information Container discharge information Container stowage info Container supply info
Custom documents
Estimated delivery date
Inland routing possibilities Route information
Stowage & load plans
Transport costs A.1 Synchronetwork transport function A.2 Container handling function A.3 Hinterland function idef0 Core synchronetwork functions
31
5.1.2 Function A.1 – Synchronetwork transport function
This section will discuss Function A.1, ‘Synchronetwork transport function’, where its hierarchy of sub-functions is depicted in Figure 8. The function does all decision making with regard to the synchronetwork trade-off. Where the function itself is decomposed by function A.11, ‘Container selection function’, function A.12, ‘Spot market pricing function’, A.13, ‘Discharging port selection’, and A.14, ‘Administrative function’. First of all the container selection function deals with selecting the containers which are of interest for the synchronetwork transport, based on the containers attributes. The spot market pricing function looks at the available routes for a container based on real-time information, and gives the transportation costs of these alternative routes. The discharging port selection function is subdivided into the ‘Handling cost determination function’ and the ‘Transport matching function’. Based on the information gathered from the container selection function, and spot market pricing function, there is determined whether it would be of interest to change the containers port of destination. The administrative function deals with all administrative tasks in relation to selecting a different inland network.
Figure 8 - Hierarchy 'Synchronetwork transport function'
Figure 9 depicts the interactions between these sub-functions. First of all, information with regard to the supply of the container, as well as how it is stowed is entering the container selection function. Here the containers which qualify for the system are determined, as well as the delivery window and stowage place. The inland routing possibilities enter the spot market pricing function, which determines the availability and prices for each alternative route, based on mode availability. In the transportation network selection function, is for each container determined whether it would be interesting to distribute it via a different inland network. This will be based upon the costs of alternative routes, the container handling, and whether the delivery window of the container is met. The output of this function gives the transportation costs, the estimated delivery date, and specific route information. The route information
A.1 Synchronetwork transport function Function A.11 Container selection function Function A.12 Spot market pricing function Function A.13 Discharging port selection function Function A.131 Handling cost determination Function A.132 Transport matching Function A.14 Administrative function Function hier Synchronetwork transport function
32 is shared with the administrative function, which will give certain changes to the stowage and load plans, as well as to custom documents.
Figure 9 - Interactions 'Synchronetwork transport function'
Figure 10 depicts the interactions between the sub-functions of the ‘discharging port selection function’, where the trade-off is made in the following way. First of all, all containers which qualify for the synchronetwork decision making enter the system. Based on the container’s location, the handling costs can be determined for each route possibility. In the transport matching function, there is made a trade-off between the total handling costs per route, and the costs savings which could be realized when following a certain route. Also the delivery window of a container is taken into account in determining whether a specific route is feasible with regard to in-time delivery. After the transportation network decision has been made, the route’s transport costs, information, and estimated delivery date can be given.
Route information Alternative routing costs
Container delivery window Container location Container stowage info Container supply info
Custom documents Estimated delivery date Inland routing possibilities
Qualified containers
Route information
Stowage & load plans Transport costs A.11 Container selection function A.12 Spot market pricing function A.13 Discharging port selection function A.14 Administrative function idef0 Synchronetwork transport function
33 Figure 10 - Interactions 'Discharging port selection function'
5.1.3 Function A.2 – Container handling function
Figure 11 shows the hierarchical order of function A.2, ‘Container handling function’, which is sub-divided into two different funtions. Namely, A.21, ‘Restowage function’, and A.22, ‘Terminal information function’. The restowage function deals with potential container handling which needs to take place when a container is assigned to a different port of destination. The terminal information function deals with all information streams regarding the handling processes required for discharging or restowing a container.
Figure 11 - Hierarchy 'Container handling function'
Figure 12 depicts the interactions between the two sub-functions of the container handling function. The stowage and load plans, with the proposed changes enter the container handling function. Based on the information given, container handling information is passed to the restowage function. This function
Alternative routing costs
Container delivery window Container location
Estimated delivery date Handling cost per route
Inland routing possibilities Qualified containers Route information Transport costs A.131 Handling cost determination A.132 Transport matching idef0 Discharging port selection function
University Edition - For Academic Use Only Date: June 23, 2019
A.2 Container handling function Function A.22 Restowage function Function A.23 Terminal information func... Function hier Container handling function
34 deals with the needed re-positioning of containers in case a new inland transportation network is selected. The information with regard to re-allocation is send back to the terminal information function, where container discharge information is generated and send further.
Figure 12 - Interactions 'Container handling function'
5.1.4 Function A.3 – Hinterland function
Figure 13 shows the hierarchical order of function A.3, ‘Hinterland function’, which is sub-divided into two different functions. Namely, A.31, ‘Inland routings function’, and A.32, ‘Route information function’. The inland routings function determines all inland transport routes available for a container. Where the route information function deals with all information available with regard to the route of a container. Which is used by the spot market function in order to determine the inland transport prices of alternative routes/modes. Function A.32, ‘Route information function’, is further sub-divided into the ‘Inland transport cancellation function’, and the ‘Communication consignee function’.
Container discharge information Container handling information Re-allocation information
Stowage & load plans
A.21 Restowage function A.22 Terminal information function idef0 Container handling function
35 Figure 13 - Hierarchy 'Hinterland function'
Figure 14 and 15 depict the interactions of the different sub-functions. First of all, container supply information is entering the inland routings function. Based on the information with regard to the container’s end destination, information with regard to the possible inland routings can be given. This information is used by the synchronetwork transport function in order to determine the market prices for these routing possibilities. A container’s route information and estimated delivery date, which is generated by the synchronetwork transport function, is bundled in the route information function. This information is of importance for the container’s consignee. When zooming into the route information function, route information enters the function. Based on this information there is determined whether pre-arranged inland transport needs to be cancelled. Based on the route information and estimated delivery date, the consignee information is generated.
A.3 Hinterland function Function A.31 Inland routings function Function A.32 Route information func... Function A.321 Inland transport cancellation func... Function A.322 Communication consignee funct... Function hier Hinterland function
36 Figure 14 - Interacions 'Hinterland function'
Figure 15 - Interactions 'Route information function'
Consignee information Container supply info
Estimated delivery date
Inland routing possibilities Route information A.32
Route information function A.31 Inland routings function idef0 Hinterland function
University Edition - For Academic Use Only Date: June 23, 2019
Consignee information Estimated delivery date
Route information Transport cancellation A.321 Inland transport cancellation function A.322 Communication consignee function idef0 Route information function
37
5.2 Synchronetwork’s economic benefits
In this paragraph, several scenarios are discussed with regard to the stowage locations of targeted containers. For the targeted containers there is assumed that they comply with the delivery window in case of being re-allocated. The red blocks in Figure 16 represent containers which block the (green) targeted containers, and therefore need to be restowed. In order to determine the potential savings for each scenario, containers can be dropped off in Rotterdam or Hamburg, which are both ports located in the Le-Havre – Hamburg route. Van Riessen et al. (2014) state that inland transportation pricing is based on the mode of transportation, as well as the distance that needs to be travelled. The costs per container for truck and train transportation will be calculated by the use of Tran (2011). Where truck costs are determined by a fixed cost of 40 euros, plus 1,2 euro times the amount of kilometres to its destination. And train costs by a fixed cost of 70 euros, plus 0,5 euro times the amount of kilometres to its destination. The distance between a port and a destination are determined by the use of google maps, and for simplicity distance is regarded similar for truck and train. The costs which need to be made for the restowage of one container are based on the information gathered during the interviews, where 200 euro per restowage was estimated. The savings are based upon the case where a container is transported with a different transport mode via the network of the other port. Since train transportation does not transport containers to its final inland destination, last-mile transportation will be needed for train transportation. After making a synchronetwork decision, certain administrative acts will need to be performed with regard to customs and stowage plans. Experts stated that these costs will be of no significance, and are therefore not taken into consideration. In order to calculate these cost savings, the following equation is formulated:
𝑆𝑦𝑛𝑐ℎ𝑟𝑜𝑛𝑒𝑡𝑤𝑜𝑟𝑘 𝑐𝑜𝑠𝑡 𝑠𝑎𝑣𝑖𝑛𝑔𝑠 =
𝐺 ∗ ((40 + 1,2 ∗ 𝑋) − (70 + 0,5 ∗ 𝑌)) − ((200 ∗ 𝐶) + 𝐺 ∗ (40 + 1,2 ∗ 𝑅))
Where,
G = Amount of targeted containers
X = Distance between port and inland destination from port-of-origin Y = Distance between port and inland destination from non-original port C = Amount of containers that need to be restowed
38 For the calculations a radius of 50 kilometers from a container’s inland unloading terminal, to its final inland destination is used in case the container is transported by train. So in order to determine the savings, the costs of transportation from Rotterdam (R) by train, compared to Hamburg (H) by truck are used. And the costs for transportation from Hamburg by train, compared to Rotterdam by truck. After determining the potential cost savings, the handling costs with regard to restowage, as well as the last mile delivery costs, are subtracted from the savings. Klink & Berg (1998) state that intermodal transportation becomes more attractive over longer distances, so three potential inland destinations are looked at in order to see the effect of distance on the concept’s performance. These inland destinations are Dortmund, Frankfurt, and Milan, which represent close, mid, and far distance locations.
5.2.1 Scenario 1 – Small selection / little restowage
Figure 16 - Scenario 1
In this scenario only little restowage is needed (1 container) in order to reach the two targeted containers. Table 2, 3, and 4 give the potential cost savings with regard to each destination.
Table 2 - Scenario 1 Dortmund
Table 3 - Scenario 1 Frankfurt
Table 4 - Scenario 1 Milan Dortmund
Amount Costs Benefits R Benefits H Benefits R Benefits H Distance in km: Rotterdam 268 Hamburg 344 Restowage moves 1 200 Cost for restowage 200
Potential containers 2 497.6 239.2 182.4 -182.4 Potential savings 40 + 1,2 * km vs. 70 + 0.5 * km
Hamburg truck 452.8 Rotterdam truck 361.6
Total benefits train instead of truck 97.6 -160.8 Hamburg rail 242 Rotterdam rail 204
Total benefits truck vs. Truck -17.6 -382.4 Last-mile in km 50 Last-mile cost 100
Frankfurt
Amount Costs Benefits R Benefits H Benefits R Benefits H Distance in km: Rotterdam 453 Hamburg 492
Restowage moves 1 200 Cost for restowage 200
Potential containers 2 667.8 535.2 93.6 -93.6 Potential savings 40 + 1,2 * km vs. 70 + 0.5 * km
Hamburg truck 630.4 Rotterdam truck 583.6
Total benefits train instead of truck 267.8 135.2 Hamburg rail 316 Rotterdam rail 296.5
Total benefits truck vs. Truck -106.4 -293.6 Last-mile in km 50 Last-mile cost 100
Milan
Amount Costs Benefits R Benefits H Benefits R Benefits H Distance in km: Rotterdam 1043 Hamburg 1110 Restowage moves 1 200 Cost for restowage 200
Potential containers 2 1561 1333.2 160.8 -160.8 Potential savings 40 + 1,2 * km vs. 70 + 0.5 * km
Hamburg truck 1372 Rotterdam truck 1291.6
Total benefits train instead of truck 1161 933.2 Hamburg rail 625 Rotterdam rail 591.5
