Yuquan Du

Modeling the Impacts of Tides and the Virtual Arrival Policy in Berth Allocation

To quantify the impacts of tides on seaside operations in container ports, this study reformulates the berth allocation problem by modeling their impacts on the entrance/exit of vessels into/from ports. Furthermore, to mitigate the tidal impacts, we borrow the so-called virtual arrival policy, whose potential for reducing bunker fuel consumption and vessel emissions is widely recognized by the shipping industry, and accordingly retrofit the berth allocation model. In the latter model, the state-of-the-art technique of second-order cone programming is adopted to handle the nonlinear intractability involved. We conduct extensive numerical experiments to evaluate the impacts of tides on the seaside operations in a tidal container port, and also to verify the competence of the virtual arrival policy in delivering win-win economic and environmental benefits for both the port and shipping lines. It is also intriguing to observe that the virtual arrival policy would be an applicable substitute for the costly approach of deepening the navigation channel in a tidal port.

A feedback procedure for robust berth allocation with stochastic vessel delays

A feedback procedure is introduced to the robust berth allocation problem with stochastic vessel delays. This procedure emphasizes the interrelationship between the planning stage and the real-time stage, and employs the cost of schedule changes in the real-time stage to instruct time buffer adjustment for each vessel. Experimental results show that this procedure performs well in different vessel delay patterns. Frequency distribution diagrams of time buffers obtained from this procedure can support the decision making of terminal planners.

Mathematically calculating the transit time of cargo through a liner shipping network with various trans-shipment policies

ABSTRACT This paper derives the mathematical expressions for the transit time of cargo through a liner shipping network. Main efforts are devoted to deriving the calculation expressions of the connection time of cargo during trans-shipment. For the forward and many-to-one trans-shipment policies, we conduct a minor correction towards the expressions in existing studies to improve the completeness. Meanwhile, we propose an alternative but more straightforward calculation method for connection time which bypasses the complicated inductive argument in existing studies. Then we introduce two new trans-shipment policies: backward trans-shipment and one-to-many trans-shipment, and mathematically calculate the corresponding connection times. Numerical experiments also deliver some managerial insights into the effectiveness of backward trans-shipment in transit time control.

Budgeting Fuel Consumption of Container Ship over Round-Trip Voyage Through Robust Optimization

A proposed practical fuel budget problem aims to determine a group of bunker fuel budget values for a liner container ship over a round-trip voyage under uncertainties caused by severe weather conditions. According to research collaboration with a global container shipping line in Singapore, the proposed problem holds a kernel position in the ship fuel efficiency management programs advocated by container shipping lines because of the downward pressure of soaring bunker prices. The synergetic influence of sailing speed and weather conditions on ship fuel consumption rate was considered when the bunker fuel budget of a ship over a round-trip voyage was estimated. To address the adverse random perturbation of fuel consumption rate under severe weather conditions, state-of-the-art robust optimization techniques were employed, and a robust optimization model for the fuel budget problem was developed. The developed model can be dualized into a mixed-integer linear programming model that may be solved by commercial optimization solvers. However, algorithmic findings in the field of robust optimization provided a polynomial time solution algorithm, and it was retrofitted to accommodate the proposed ship fuel budget problem. The case study of an Asia–Europe service demonstrates the computational performance of the proposed solution algorithm and the competence of the proposed robust optimization model to produce fuel budget values at different levels of conservatism possessed by the fuel efficiency specialists in container shipping lines.

Bunker procurement planning for container liner shipping companies: Multistage stochastic programming approach

This paper investigates the bunker procurement planning (BPP) problem arising for a container liner shipping company that plans to purchase bunker from both bunker futures contracts and the spot market to hedge the risk in fluctuation of and increases in bunker prices. A multistage bunker procurement decision process for the BPP problem is developed to determine the monthly bunker procurement. The process allows the shipping company to sign bunker futures contracts in the first stage and to rebalance them in the subsequent stages. By assuming the stochasticity of bunker spot price, the BPP problem is formulated as a mean-variance minimization stochastic programming model. An approximation solution method for solving this model is designed by integrating random variable sampling technique, scenario tree generation, and quadratic programming approximation. Finally, numerical experiments demonstrate that bunker procurement risk can be effectively hedged with the proposed method. This study provides a useful decision tool for container liner shipping companies to use when planning bunker procurement.

Dynamics and interdependencies among different shipping freight markets

ABSTRACT An appropriate description of freight rate behaviors is important to maritime forecasting and portfolio diversification in shipping freight markets. We employ general autoregressive conditional heteroscedasticity-copula models to capture the dynamics and interdependencies among shipping freight rates. Using weekly data from 5 January 2002 to 24 March 2018, our main findings are first, Granger causality tests confirm the presence of one-way causality running from the dry bulk and the clean tanker freight rate returns to the container and the dirty tanker freight rate returns, respectively. Second, volatility persistence exists in individual shipping freight market and, in particular, it is much less persistent in the clean tanker freight market. Third, nonlinear dynamic interdependencies among freight rate returns are captured by performing time-varying copulas. The results not only deepen our understanding of freight rate behaviors but also offer new insights into portfolio diversification and risk management in the shipping freight markets.

Integrating fuel consumption and vessel emissions into berth allocation

This paper focuses on the berth allocation problem in the container terminal, considering fuel consumption and vessel emissions. By regarding the arrival times of vessels as decision variables and introducing the relationship between the fuel consumption rate and the sailing speed in the form of general power functions, we obtain a more elaborate model on the berth allocation, considering fuel consumption, than ever before. In order to overcome the computational complexity, we cast the original mixed integer nonlinear programming model as a mixed integer second order cone programming model. Furthermore, based on the fuel consumption calculation, we conduct the vessel emission calculation with the widely-used emission factors. Numerical experiments show that the proposed model can significantly reduce the fuel consumption and the vessel emissions in sailing periods, without lowering the service level of the terminal. This effort on maritime emissions is consistent with the environmental concerns of the international maritime organization.