Economic speed and repositioning of a tramp ship under uncertain fuel consumption and future profit potential

Abstract

Given the nature of sailing and the unique features of tramp shipping, the decision-maker who is considered to be the carrier in a shipping contract may face uncertainties from the natural environment, ports, shipping markets, and the counter-party. This impacts decisions about not only the journey itself, i.e., job acceptance, shipping route, schedule, and selection of terms, but also future operations after the vessel terminates at the destination port. The topic of economic decision-making of speed for tramp ships has been addressed in existing research from the following perspectives: deterministic framework (Ronen, 1982; Ge et al., 2021); uncertain weather performance (also known as weather routing and scheduling problems in the marine engineering community) (Zis et al., 2020); and other uncertainties (Hwang et al., 2008; Agra et al., 2013; Lindstad et al., 2013; Christiansen and Fagerholt, 2014). This thesis provides insights on both the development of fundamental optimisation models, including mean-risk optimisation models and Markov Decision Processes (MDPs), and practical analysis for decision-making about job acceptance, route and speed scheduling, and the selection of freight payment terms. In Chapter 3, we develop a new perspective on the problem of economic (average) ship speed by considering the impact of the decision maker's risk attitude. This impacts decisions about both speed and job acceptance. A ship is used to transport bulk cargoes in the spot market. A job consists of moving a cargo from a port A to a port B. Whether a ship owner can accept a job is determined by (i) the profitability of this job, and (ii) the commercial value of having its ship in port B when the job is finished. The time needed to travel from A to B affects both (i) and (ii). We consider that the decision maker wishes to maximise the Net Present Value (NPV) of the ship under uncertainty. This uncertainty is associated with the fuel consumption on the journey and the future profit potential of the ship at the next port. Underlying factors for these risks include adverse fuel consumption rates, and randomness in freight markets. We develop mean-risk optimisation models based on either long-term or short-term risk perspectives, justify why this distinction is worthwhile to consider, and introduce stochastic programming methods to solve the set of models. Numerical experiments illustrate the approach and show the sensitivity of the optimal strategy to context parameters and risk attitude of the decision maker. The mean-risk speed optimisation models can be extended to account for risk from different sources, e.g. failure-to-pay. In Chapter 4, we develop a method of dynamic stochastic programming for solving a ship routing and scheduling problem when fuel consumption as a function of speed depends on the location of the ship and time. More specifically, a framework of a Markov Decision Process (MDP) incorporates the stochasticities beyond and after terminating at the destination port, as well as the information updating through the decision process. In this study, the decision-maker expects to maximise the long-term profitability, which includes not only the profit obtainable from the current journey but also the profitability after termination. We employ the approach of Net Present Value (NPV) and exploit the Future Profit Potential (FPP) to represent profitability after completing the current journey. The model is established in 3D states that include the spatial and temporal constituents of the vessel and solved by the value iteration algorithm. Subsequently, a simulation-enhanced value iteration (SEVI) is proposed to generate the distribution profile of NPV in the short- and long-term for decision-makers with a variety of risk attitudes. Numerical experiments show the methods proposed in this paper generate a higher NPV under a wide range of scenarios under risk. Experiments in reference to alternative delivery time-windows offer insights into how to secure an ideal delivery time-window before reaching agreement with due consideration of risk attitudes and long-term profitability. In Chapter 5, we argue for the consideration of non-payment risk into the decision-making problems for the carrier in tramp shipping. We employ a Net Present Value (NPV) model to generally describe payment structures under a variety of freight payment terms, including Freight Prepaid (FP) and Freight Collect (FC), and shipment terms, including Free-on-Board Origin (FOB-O), Free-on-Board Destination (FOB-D), Cost and Freight (CFR), Cost, Insurance, and Freight (CIF). We demonstrate the cash-flows are symmetric when all parties have extended trust when conducting business activities with each other, which means there is no non-payment risk with freight charges. While the trust among all parties is weaker than extended trust, i.e., basic trust or guarded trust, additional monitoring mechanisms are required to assure all parties are liable to comply with the contract. Letter of Credit (LC), as a financial instrument that has been widely used in international trade, is discussed in this paper, especially for the transaction for freight charges. Variations in Letters of Credit are addressed in the payment structure. Computational results reveal that when the unit freight rate is determined, Red Clause Letters of Credit (RCLC) are more advantageous than other types of LC, i.e., Irrevocable Confirmed Letters of Credit (ICLC), and Letters of Credit at sight (LC at sight), especially when the carrier has inadequate liquidity cash flows. Whereas LC at sight first-order stochastic dominants RCLC under specific conditions.

More information

Identifiers

ORCID for Yuanming Song: orcid.org/0000-0003-2847-3013

ORCID for Patrick Beullens: orcid.org/0000-0001-6156-3550

ORCID for Dominic Hudson: orcid.org/0000-0002-2012-6255

Catalogue record

Export record