PORTS PCP: SIMULATING THE SHIP-ANCHORAGE CONNECTION TO REDUCE DEMURRAGE COSTS – PART 1

Several researchers have focused their work on studying the application of sophisticated modeling techniques to a port environment and the complex relationships between cost and service levels that may eventually arise. Specifically in relation to the operation of the ship-anchor connection (SBL). ship-berth link), it can be argued that whatever the type of cargo handled in ports, ship operators do not appreciate queue management because of berth congestion (Dasgupta and Ghosh, 2000). In fact, ship operators that do not receive certain guarantees of berth availability may seek alternative ports to maintain high levels of productivity (Luo and Grigalunas, 2003). As a direct consequence, ports were forced to consider the quality of the service offered and also the total costs of demurrage (overstay).

If port authorities can plan and control SBL operations on a daily basis in the short term, given the characteristics of the ships and the market segments they serve in the medium/long term, they will be able to adequately plan future investments in the expansion of berths and they will also be better prepared to deal with increasing competition (Ho and Ho, 2006). Within this context, the simulation of port operations may be an option to address these issues, as it contributes to assessing port performance and generating different scenarios to help in decision-making (Duinkerken et al., 2006).

The concept of simulation is not new. Simulation as a research tool dates back to the 1960s and has been applied to different areas of logistics research, such as the design and operation of transportation systems, the layout and handling of warehouse materials and the determination of ordering policies for inventory systems (Casaca, 2005).

There is a growing number of studies dealing with SBL planning (Imai et al., 2001 and 2005). SBL refers to the interface between the land and water parts. Traditionally, the SBL planning problem comprises the allocation of ships arriving at mooring positions, as well as the scheduling of quay cranes, which play an important role in the management of port operations (Meisel and Bierwirth, 2008). Simulation models have been widely used in SBL problem planning and analysis (Dragovic et al., 2005).

In this work, we focus on the analysis of vessel waiting statistics and on the total demurrage costs under different berth allocation norms and queue priorities, through simulation models developed in the Arena software. The model is supplied with data from a confidential case study about a private Brazilian port, in the state of São Paulo. The results are analyzed in terms of the most suitable combination of berth allocation norm and queuing priorities for a given set of critical levels of queuing waiting times (after which overstaying costs are incurred) and different coefficients of costs of demurrage between large and small ships.

LITERATURE REVIEW

Casaca (2005) presented a comprehensive structure of port industry operations, describing in detail its three main subsystems: the ship-side interface or berth area, the container yard, and road and rail access gates. Its framework makes it clear that port operations are complex by nature and therefore require sophisticated modeling techniques such as simulation, generic algorithms and non-linear programming to help port authorities in different aspects of decision-making. The structure presented by Casaca is also useful for mapping and organizing the research found in the literature according to its main motivators, the modeling technique adopted, the subsystem studied and the main decisions dealt with.

For example, as most shipping companies have started to operate large container ships in recent years, several authors have used simulation to analyze their impact on the container yard under different situations. Chang (2005) divided a container terminal into three subsystems similarly to Casaca and modeled different operation patterns involving mooring operations, in order to test possible queuing scenarios in the container yard. Similarly, Tu and Chang (2006) used simulation software to build different models of docking operations and examined possible scenarios in the container yard. Both surveys found that ship-side operations could be largely responsible for container yard delays.

Regarding gates, Parola and Sciomachen (2005) presented a discrete event simulation model to analyze how to face the impact of growth in maritime traffic on infrastructure on land. More precisely, the authors studied the impact of the saturation level on the railway lines and on the level of congestion at the truck access gates. In turn, Kim et al. (2003) suggested a dynamic programming model for truck arrivals. Several sequential rules were also compared through simulation. The authors found that the shortest processing time rule showed a robust and high level of performance in several situations.

Container handling and inter-terminal transport systems may also be subject to analysis. For example, Duinkerken et al. (2006) presented a comparison between three transport systems for inland transport of containers between terminals. Ottjes et al. (2006) proposed a generic simulation model framework for the design and evaluation of multiterminal systems for container handling. In both studies, the experiments carried out gave a better insight into the importance of different characteristics of conveyor systems and their interaction with handling equipment.

However, regardless of the nature of the port subsystem under analysis and the modeling technique employed, one thing is clear: the competitiveness of a port is especially measured in terms of an adequate level of service offered to ship operators or users (Legato and Mazza , 2001). Thus, a central goal for port authorities is to reduce the waiting time of ships, from the moment they arrive at the port until the moment of departure, through better management of current resources, which certainly requires a large expenditure of capital and long payback period (Ho and Ho, 2006). In this sense, a more attentive consideration of the ship-side interface, and more particularly of the SBL, is considered necessary. This link is not only responsible for a substantial part of the investment required to build a port, but also for the total waiting time in the queue until the start of (un)loading operations.

SBL operation has been variously referred to as berth planning system (Legato and Mazza, 2001), berth allocation planning (Nishimura et al., 2001), berth allocation problem (Meisel and Bierwirth, 2008) and anchorage-crane operations (Canonaco et al., 2008). In general terms, the main task of the SBL operation is to allocate a limited number of berths between incoming ships. Choosing to moor one ship rather than another at a specific anchorage can result in a very long distance from the point where your containers are located in the yard, generating cross-effects in terms of delays, not only in container yard operations, but also in the docking queue (Meisel and Bierwirth, 2008).

Another common feature of SBL operation is the need to address resource limitations and physical constraints. The limited number of berths and quay cranes restrict the service capacity of ports and terminals, often leading to a trade-off between investments in fixed assets and the total costs of overstaying. In these cases, the use of simulation as a planning tool has been increasingly important to find a balance between queuing priorities and the possibility of postponing these investments.

Dragovic et al. (2005), who reported on several different simulation models in relation to port operations, particularly examined the impact of introducing priority, for certain classes of ships, on SBL performance. One of their simulation models indicated that assigning priority to smaller ships would lead to an improvement in key operational parameters such as the average number of ships in the queue and the average time a ship spends in the queue. One can easily see the crucial role that the time spent in the queue plays in reducing the total waiting time of ships from the moment they arrive at the port until the moment of departure. The path to an agile port environment must involve some fundamental aspects of operations management (Casaca, 2005), such as, for example, a detailed study on queue priority rules and berth allocation.

As highlighted by Asperen et al. (2003), priority rules are expected to reduce waiting costs for high-priority ships. In his study, a simple scheme with two priority classes – high and low – was considered, in which large ships had high priority, and small ships had low priority. Applying priority rules has reduced the percentage of high priority ships and increased the percentage of low priority ships that have to wait. However, the question of whether total demurrage costs are reduced by applying priority rules, or to what extent, depends on how much more expensive an idle high-priority ship is than a low-priority ship.

This remains to be tested and, according to Dragovic et al., as cost is an essential measure in choosing alternative strategies to the SBL problem, further research is needed to incorporate a cost analysis in order to identify the combination most appropriate set of berth allocation norms and queue priorities in a given context.

In this work, we consider the simulation case study in a private Brazilian container port as a starting point to address these issues. As the port is relatively small, with only two berths, some simplifications have been made here to the SBL model in order to increase the focus on suiting different berth allocation norms and queue priorities.

Different critical levels for waiting times in queues and penalties for overstaying are analyzed in an operation in which the choice of berthing a ship instead of another in a specific anchorage is considered to have no impact on the distance in relation to the point of location of your containers in the yard. Another simplification relates to the number of wharf lifts and cranes: their impact on the operation is considered to be integrated into the average processing times for each ship, regardless of the berth used. Finally, it was also considered that the rest of the port's resources, in addition to the scope of the SBL operation modeled in Arena, do not affect the time spent in the queue for each ship.

OBJECTIVES

Based on the literature review, the following questions were defined for this simulation study. The first concerns the quantification of the impacts of berth allocation and queuing priorities on the time a ship spends in the queue. It is given as follows:

1 – What is the impact of different berth allocation rules and queue priorities on waiting times for the port system as a whole and for each of the ships that periodically visit the port?

In this study, four different berth allocation norms were considered:

• Dedicated berths by ship type. That is, depending on the size of the ship, one berth caters exclusively to small ships and the other to large ships;
• Single file dispatches ships to the first available berth. In this case, when all berths serve both types of ships (small and large), they are kept in only one row before being directed to the first available berth;
• Ships allocated to the anchorage with the smallest queue size. According to this standard, ships enter the queue with the smallest size at any given time, thus implying that all berths cater for all types of ships. Unlike the last rule, there is no single queue that detains ships until a berth becomes available;
• Ships allocated to the anchorage with the shortest queue time. Rule similar to the last one, with the exception of the fact that ships enter the queue with the shortest expected time to start berthing at a given moment, based on the sum of the expected time for processing of ships that are already waiting in the queue.

Regarding queuing priorities, eight different disciplines were considered within each mooring allocation norm (Hansen, 1972; Silberholz et al., 1991). These disciplines are considered to decide, at the time a berth becomes available, which ship in the queue will be serviced next. These are given below:

• Longer processing time.According to this priority, the ship with the longest processing time, when a berth becomes available, is served first. This priority is the opposite of Shorter processing time;
• Highest number of ship arrivals per year.According to this priority, the ship with the highest number of visits scheduled per year is served first. This priority is the opposite of Lower number of ship arrivals per year;
• PEPS(First in, first out) and LIFOs (Last in, first out), two classic and well-known queuing disciplines (see eg Nahmias, 2001);
• Larger ship.According to this priority, the largest ship is served first when a berth is empty. This priority is the opposite of Smaller ship size.

The possible alternatives for the operation of SBL are illustrated in Figure 1 through different flowcharts. These were modeled in Arena 8.0, a well-established simulation tool for discrete events (Chung, 2004). They basically indicate the same sequence of activities and decisions, differing only in the factors being tested. In general terms, ship care begins with its arrival at port. Depending on congestion and the priority assigned to the incoming ship, it may have to wait until a berth becomes available. After berthing, the containers are unloaded and/or loaded onto the ship. Finally, when the service is completed, the ship leaves the port.

(1) Dedicated berths per vessel size

(2) Deploys ships to the first available berth

(3/4) Ships allocated to an anchorage with smaller size or queue time (Q1 and Q2)

Figure 1: Flowcharts

BIBLIOGRAPHY

Asperen EV, Dekker R, Polman M, and Arons HS. 2003: Allocation of ships in a port simulation. Proceedings of the 15th European Simulation Symposium ISBN: 3-936150-29-X (CD).

Canonaco, P., Legato, P., Mazza, RM and Musmanno, R.. 2008: A queuing network model for the management of berth crane operations. Computers & Operations Research 35: 2432-2446.

Casaca, ACP. 2005: Simulation and the lean port environment. Maritime Economics & Logistics 7: 262-280.

Chang, YF. 2005: Analysis of operations at the Kaohsiung port new mega container terminal. The Business Review 4: 123-128.

Chung, CA. 2004: Simulation modeling handbook: a practical approach. CRC Press: Boca Raton.

Dasgupta, A. and Ghosh, M.. 2000: Inducing performance in a queue via prices: the case of a riverine port. Management Science 46: 1466-1484.

Dragovic, B., Park, NK, Radmilovic, Z. and Maras, V.. 2005: Simulation modeling of ship-berth link with priority service. Maritime Economics & Logistics 7: 316-355.

Duinkerken MB, Dekker R., Kurstjens STGL, Ottjes JA and Dellaert NP. 2006: Comparing transportation systems for inter-terminal transport at the Maasvlakte container terminals. OR Spectrum 28: 469-493.

Hansen, JB. 1972: Optimizing ports through computer simulation: sensitivity analysis of relevant parameters. Operations Research Quarterly 23: 519-530.

Ho, MW and Ho, KH. 2006: Risk management in large physical infrastructure investments: the context of seaport infrastructure development and investment. Maritime Economics & Logistics 8: 140-168.

Imai, A., Nishimura, E. and Papadimitrou, S.. 2001: The dynamic berth allocation problem for a container port. Transportation Research Part B 35: 401-417.

Imai, A., Sun, X., Nishimura and Papadimitrou, S.. 2005: Berth allocation in a container port: using a continuous location space approach. Transportation Research part B 39: 199-221.

Kim, KH, Lee, KM and Hwang, H.. 2003: Sequencing delivery and receiving operations for yard cranes in port container terminals. International Journal of Production Economics 84: 283-292.

Legato, P. and Mazza, RM. 2001: Berth planning and resources optimization at a container terminal via discrete event simulation. European Journal of Operational Research 133: 537–547.

Luo, M. and Grigalunas, T.. 2003: A spatial-economic multimodal transportation simulation model for US coastal container ports. Maritime Economics & Logistics 5: 158-178.

Meisel, F. and Bierwirth, C.. 2008: Heuristics for the integration of crane productivity in the berth allocation problem. Transportation Research Part E doi: 10.1016/j.tre.2008.03.001.

Nahmias, S.. 2001: Production and operations analysis. McGraw-Hill: New York.

Nishimura, E., Imai, A. and Papadimitriou, S.. 2001: Berth allocation planning in the public berth system by genetic algorithms. European Journal of Operational Research 131: 282-292.

Ottjes, JA, Veeke, HPM, Duinkerken, MB, Rijsenbrij, JC and Lodewijks, G.. 2006: Simulation of a multiterminal system for container handling. OR Spectrum 28: 447-468.

Parola, F. and Sciomachen, A.. 2005: Intermodal container flows in a port system network: analysis of possible growths via simulation models. International Journal of Production Economics 97: 75-88.

Silberholz MB, Golden BL and Baker EK. 1991: Using simulation to study the impact of work rules on productivity at marine container terminals. Computer & Operations Research 18: 433-452.

Tahar, MR and Hussain, K.. 2000: Simulation and analysis for the Kelang Container Terminal operations. Logistics Information Management 13: 14–20.

You, YP and Chang, YF. 2006: Analyzes of operations of ditch container wharf and container yard. The Journal of the American Academy of Business 9: 139-146.