A comprehensive annual delivery program for upstream liquefied natural gas supply chain

Highlights

• We develop a comprehensive model for LNG ADP problem.

• We model several practical aspects of ADP problem overlooked in the literature.

• We provide insights on how additional aspects of the model affect the ADP.

• We develop a novel solution that affectively solves realistic size problems.

• We establish how various problem parameters affect solution quality.

Abstract

Developing a cost-effective annual delivery program (ADP) is a challenging task for liquefied natural gas (LNG) suppliers, especially for LNG supply chains with large number of vessels and customers. Given significant operational costs in LNG delivery operations, cost-effective ADPs can yield substantial savings, adding up to millions. Providing an extensive account of supply chain operations and contractual terms, this paper aims to consider a realistic ADP problem faced by large LNG suppliers; suggest alternative delivery options, such as split-delivery; and propose an efficient heuristic solution which outperforms commercial optimizers. The comprehensive numerical study in this research demonstrates that contrary to the common belief in practice, split-delivery may generate substantial cost reductions in LNG supply chains.

Introduction

Natural gas (NG) is the third largest energy source satisfying approximately 25 percent of the world’s energy demand (BP Department, 2014b). Pipelines are the major delivery method in global NG trade. Although pipeline delivery is highly cost efficient over short distances, it is not technically feasible and economically viable when the supplier and customers are geographically spread. In such cases, NG is transported in the form of liquefied natural gas (LNG), utilizing special ocean-going vessels over long distances. Shipping NG in the form of LNG offers gas suppliers the ability to reach distant markets.

Global LNG trade has been steadily growing for the past two decades. Its share in the global NG trade increased from 26 percent in 2000 to 32 percent in 2013 (BP Department, 2014b). This share is expected to grow to 46 percent by 2035. Similarly, LNG accounts for 10 percent of the world’s NG consumption, which is expected to increase to 15 percent by 2035, according to BP Department (2014a). This increase in LNG demand, coupled with the increasing production and delivery capacities, has led to complex global LNG supply chains. Currently, global annual liquefaction capacity is 408 billion cubic meters (BCM), which is expected to increase to 1427 BCM when the construction of new liquefaction terminals is completed (Enerdata, 2014). On the shipping side, there are currently 405 LNG vessels with a total capacity of 59.5 million cubic meters (MCM); and 132 more vessels are currently being built to increase this capacity to 81 MCM (Enerdata, 2014).

Both the production and the delivery of LNG require significant capital investments and operating expenses. Capital cost of a liquefaction plant can be as high as 600 dollars per ton per annum production while capital cost of an LNG vessel ranges between 200 and 400 million dollars. Operational costs are also significant; the cost of daily vessel charters averaging around 60,000 dollars (Tusiani & Shearer, 2007). In light of these figures, the vast majority of global LNG trade needs long-term contracts to secure intensive capital investments in the LNG supply chain and ensure income stability for LNG suppliers. Meanwhile, due to price volatility across NG markets and excess LNG production, LNG spot market has been growing very rapidly. According to Kent (2013), the share of spot sales in the LNG trade has risen from 6 percent in 2000 to 25 percent in 2012.

In order to survive in such a dynamic and complex market, suppliers should carefully plan for LNG production, inventory, and delivery in a cost effective way. Planning problems arise at strategic, tactical, and operational decision levels: Strategic planning typically involves decisions regarding production and vessel investments as well as long-term contract terms and conditions. Tactical planning involves developing an integrated fulfill, inventory, and delivery plan for a period of 12-to-18 months to satisfy the long-term contracts’ requirements. This plan is called the annual delivery program (ADP). Finally, operational planning typically includes mitigations to planned or unplanned changes in the ADP such as vessel breakdowns, port unavailability, variations in travel times, and emerging spot sales opportunities.

In this paper, focusing on tactical planning, we study the problem of preparing a cost-effective ADP for a global LNG supplier. This is an integrated production, inventory, and maritime routing problem. The supplier operates a liquefaction plant, where NG is cooled to $-240$ degree Fahrenheit and condensed into liquid form, which is called LNG production. The LNG is temporarily kept in storage tanks at a port near the liquefaction facility before it is loaded on LNG vessels. The supplier owns and/or charters a fleet of heterogeneous vessels that are used for delivering LNG to customers with long term contracts. For each customer, there are several contracts throughout a certain year, each of which comes with differing levels of flexibility in terms of delivery volume and time. In most practical cases, total annual long term contractual demand is less than the production capacity. Due to highly limited production changes and lack of facilities to store large amounts of LNG in storage tanks, there will usually be some excess production. Therefore, the ADP should also include a provisional plan for selling the excess production in the spot market.

In the ADP problem, we would like to minimize the sum of operating costs, penalty costs due to deviations from contractual terms, and penalty costs of handling excess production. We develop a novel heuristic method to create a cost effective ADP in a short amount of time. We then interpret the results from a comprehensive computational study to gain managerial insights.

There are a handful of studies that attempt to develop cost-effective ADPs, such as Halvorsen-Weare and Fagerholt (2013), Rakke et al. (2011) and Stälhane et al. (2012). Our main contributions in this study are incorporating many practical aspects of upstream LNG operations and LNG contracts into the ADP preparation problem and proving managerial insights regarding these aspects. Furthermore, we propose a novel and efficient heuristic solution approach for this problem, which can be used as a practical decision support tool by the logistics planners of LNG suppliers. This heuristic consistently outperforms commercial optimizers by finding less-costly ADPs in a considerably shorter amount time.

The remainder of this paper is organized as follows: in Section 2, we present a review of the relevant literature and discuss the contribution of our work. In Section 3, we describe the mathematical model and present the formulation. In Section 4, we introduce our heuristic method to construct an ADP. Finally, we present our numerical analysis in Section 5 and conclude with model extensions and further research directions in Section 6.