Supply chain design and planning for LNG as a transportation fuel
Lopez Alvarez, Jose
DOI:
10.33612/diss.131459842
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2020
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Lopez Alvarez, J. (2020). Supply chain design and planning for LNG as a transportation fuel. University of Groningen, SOM research school. https://doi.org/10.33612/diss.131459842
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
Chapter 1
Introduction
Supply chain management decisions play an important role in the transition from fossil fuels to alternative, cleaner fuels. The outcomes of these decisions shape the design and planning of an alternative fuel network and thus determine whether the alternative fuel is distributed to its users in an efficient and cost-effective way. Nu-merous recent scientific studies have investigated various decision-making problems arising in the planning and design of the distribution networks for alternative fuels. Some studies have addressed the facility location problem of alternative fuel sta-tions, considering various characteristics such as undeveloped market demand (e.g. Capar et al. (2013)) or limited space or capacity to serve customers (e.g. Wang and Lin (2013)). Similarly, in the context of biofuels, multiple studies have investigated facility location problems for biofuel production plants (e.g. Chen and ¨Onal (2014)) and the inventory control problems that emerge when taking into consideration the different types of raw materials that could be used to produce biofuels (e.g. ˇCuˇcek et al. (2012)).
This thesis addresses decision-making problems arising in the distribution net-work for liquefied natural gas (LNG) and its biofuel counterpart, Bio-LNG. LNG is a cryogenic fuel; that is, it needs to be maintained at extremely low temperatures (namely −162◦C) to keep it in a liquid state. In the context of road freight, inland waterway and maritime transportation, LNG is a promising alternative fuel for the short and medium term. Compared with the well-established fossil fuels that are typically used in those sectors (e.g. diesel and heavy fuel oil), LNG generates lower
carbon emissions and other air pollutants such as nitrogen oxides (NOx) and sulfur
dioxide (SO2). In addition to the environmental advantages of LNG, reserves of
nat-ural gas are vast, which guarantees the supply of the fuel for at least several decades (Arteconi and Polonara, 2013).
The merits of LNG as a transportation fuel have not been ignored. The 2014/94/EU Directive of the European Parliament, in which a framework for the deployment of the infrastructure for alternative fuels in the European Union is sketched, presents a number of provisions regarding deployment of the distribution network for LNG. These are aimed at ensuring that a refueling infrastructure for trucks and ships run-ning on LNG is deployed across the Trans-European Transport Network (TEN-T) by the end of 2030. As a result, the refueling infrastructure would allow trucks and ships running on LNG to move across the TEN-T network without running out of fuel.
The typical distribution network for LNG as a fuel is an extension of the global LNG supply chain, where the commodity is traded in large volumes among produ-cing and importing regions. Upon production, the LNG is shipped in LNG carriers from export terminals (e.g. the Australia Pacific LNG terminal) to large-scale import terminals (e.g. the Gate terminal in Rotterdam). All the imported LNG is either re-gasified and fed into the natural gas grid or stored as a liquid to be later distributed to consumers. For those LNG refueling facilities/end-users that are close to an im-port terminal, replenishment of fuel can be done directly from the imim-port terminal by means of specialized tanker trucks or barges. When refueling facilities/end-users are further away, satellite facilities may be needed to distribute the fuel in a cost-effective manner.
In principle, end-users can also be supplied by local producers of Bio-LNG, which is compatible with LNG as a transportation fuel. Bio-LNG is a renewable fuel that can be produced, for example, by liquefying biogas, which is a mixture of gases pro-duced from renewable sources. Current production levels of Bio-LNG are not nearly enough to sustain the demand for LNG as a fuel on its own. This is mostly due to the limited availability of biomass and the high operating costs of existing technolo-gies to generate Bio-LNG, as well as the high level of investment required for these technologies.
In order to establish a distribution network for LNG as a fuel that ensures that it is accessible to end-users at a competitive price and of the right quality, substantial
Introduction 3 challenges must be overcome (Thunnissen et al., 2016; Simmer et al., 2014). Regu-lations, standardization of procedures and financial support are among the critical challenges that need to be tackled. From a supply chain management perspective, we highlight two major challenges related to the cryogenic nature of LNG. First, both the distribution network and demand for LNG as a fuel are noticeably un-derdeveloped. Demand is expected to grow as the distribution network matures. However, such network investments are capital-intensive and risky owing to the specialized nature of the infrastructure required to store and transport LNG. As a result, infrastructure investment decisions require increased demand, which results in a chicken-and-egg problem. Second, LNG constantly boils off because of its cryo-genic nature. This boil-off not only reduces the quantity of the fuel, but also affects its quality. The use of low-quality LNG as a fuel reduces its efficiency and, in extreme cases, can permanently damage the engines of the end-users.
The consideration of these challenges in the design and planning of LNG dis-tribution networks gives rise to complex decision-making problems. First, decisions regarding the establishment of satellite and refueling facilities for LNG as a fuel must ensure that the distribution network is large enough to guarantee the accessibility of the fuel to the market. However, if the infrastructural development is greater than needed, the distribution costs of the fuel will increase and thus so will the market price. Accordingly, it is crucial to identify the distribution network design with the minimum infrastructure needed to enable the adoption of the fuel. Second, the man-agement of LNG inventories in the different facilities along the distribution network must take into consideration the quality of the fuel and its deterioration. Effect-ive decision-making regarding the management of LNG inventories can ensure that customers are supplied with LNG that meets the minimum quality requirement in a cost-effective manner.
Inspired by the above decision-making problems arising in the design and man-agement of an LNG distribution network, this thesis addresses novel operations re-search problems in the domain of location-routing, lot-sizing and inventory control problems. In each of the problems studied in this thesis, we extend the existing liter-ature by incorporating the characteristics inherent to LNG or its market that have not been taken into consideration in previous studies. From a methodological point of view, we design and propose effective solution approaches for the aforementioned problems. These approaches are derived from a thorough analysis of the problems
and their characteristics. From a practical point of view, this thesis provides various insights regarding the infrastructure development of LNG and also about the invent-ory management decisions that need to be made while taking quality into consider-ation. Some of these insights can be applied to distribution networks through which other cryogenic fuels and/or alternative fuels are distributed. Each of the problems to be studied in this thesis relates to one of the two challenges mentioned above, i.e. the infrastructure development of the LNG network and the management of LNG inventories while taking quality into consideration. Accordingly, in the following sections, we elaborate on each of the two challenges; as we do so, we also briefly introduce the research projects in this thesis.
1.1
Infrastructure development of an LNG distribution
network
In the early market phases of an alternative fuel, the infrastructure available to sup-ply the fuel to its end-users is limited. The investments required to expand the dis-tribution network are capital-intensive and involve high risks due to the uncertainty regarding the extent to which the market will adopt the alternative fuel. As a res-ult, infrastructure developers hold their investments until the market reaches a ma-ture phase. At the same time, potential end-users typically do not invest in vehicles fueled by an alternative fuel until there is a distribution network that ensures the accessibility of the fuel at a competitive price. This results in a chicken-and-egg problem, where both the demand and supply sides of the market are discouraged from adopting alternative fuels.
In the context of LNG in the European Union, the status of the LNG distribution network contrasts with the ambitions of the 2014/94/EU Directive, the idea of which is to ensure that trucks and ships running on LNG would be able to move across the TEN-T network without running out of fuel. Currently, the existing infrastructure for the distribution of LNG in the European Union is mostly concentrated in nations where LNG import terminals are located. Around 80% of all the LNG refueling sta-tions for road transport in the European Union are located in four of the countries with large-scale import terminals (namely Spain, Italy, France and the Netherlands), while the infrastructure for LNG as a fuel is largely absent in the rest of the European Union (NVGA, 2019). Furthermore, bunkering operations (i.e. refueling operations
Introduction 5 for maritime transport) are mostly available on major European ports (e.g. the ports of Rotterdam and Antwerp); bunkering LNG through the major European water-ways, such as the Danube river, is still limited. Although there are ongoing projects to further develop the infrastructure of the LNG distribution network, especially along some of the major waterways, many more development projects are needed to meet the 2014/94/EU Directive.
Infrastructure developers and decision-makers in the LNG market seek to identify the minimum infrastructure required to enable the adoption of LNG. In the context of truck operators, a key element in adopting an alternative fuel is the availability of refueling infrastructure in the area of operations. When the infrastructure available to refuel LNG is not sufficient, trucks need to make long detours to reach refueling stations, which not only increases costs but also diminishes the environmental bene-fits of LNG compared with other fossil fuels. Therefore, decision-making regarding the deployment of the refueling infrastructure should not be based solely on profit or cost, but also on the accessibility of the fuel to end-users. In a study leading up to the research in this thesis (Post et al., 2018), we developed an approach to identify the minimum infrastructure required to prevent long detours for truck operators, which would discourage the adoption of LNG as a fuel. In that research, we used the adop-tion of LNG as a transportaadop-tion fuel in the Netherlands as a case study. Interestingly, our results indicated that the minimum number of refueling facilities needed in the country was 25, which is approximately the number of LNG refueling facilities today in the Netherlands.
Currently, most replenishment of refueling facilities for both road and maritime transportation is done directly from the LNG import terminals. However, direct re-plenishment from the import terminal may not be economically effective for refuel-ing facilities located in nations without such terminals. As a result, satellite facilities could play a key role in these cases. Such facilities can exploit the economies of scale that can be achieved by replenishing their storage tanks by means of barges/vessels that transport LNG; these barges/vessels can carry several times the amount of LNG that can be transported in tanker trucks. The possibility of establishing satellite fa-cilities that can be replenished using LNG barges/vessels has been already acknow-ledged in previous research. For example, Jokinen et al. (2015) present a model to support decision-making regarding the establishment of LNG satellite facilities is the south of Finland. In that study, customers are fulfilled with dedicated trucks that
may originate at the terminal or at a satellite facility; however, a customer can only be replenished by a facility that is not more than 300 km from the customer’s location. In another related study, Bittante et al. (2018) present a fleet composition problem for small-scale LNG operation. In that problem, the LNG needs to be transported from supply ports to demand points using (potentially) various types of ships with different capacities to hold LNG.
In Chapter 2 of this thesis, we develop a model to support decision-makers of the small-scale LNG network with regard to investments in new satellite facilit-ies, bunker barges and tanker trucks. Inspired by current practice, we study a set-ting where the capacity of transportation resources is exploited by making routes in which the demand of various end-users/refueling facilities is satisfied. Addition-ally, we allow direct deliveries from the import terminal to any (or multiple) demand points, bypassing the satellite facilities—which is how most refueling facilities are currently replenished in practice. We propose an algorithm to solve the problem and present a case study inspired by the network design problem for some of the main corridors of the TEN-T network. Our results from the case study reveal that satellite terminals and bunker barges become an interesting option when demand for LNG grows and occurs farther away from the import terminal. In those situations, the large investments associated with LNG satellites and bunker barges are offset by reductions in operational costs of the LNG tanker trucks.
1.2
Management of LNG inventories considering
quality
Throughout the distribution network of LNG, the fuel is transported and stored in well-insulated storage tanks that minimize the exposure of the LNG to heat leakages. Since there is no perfect insulation, the ingress of heat to the storage tank causes con-tinuous vaporization of the fuel (Dobrota et al., 2013). This vaporization not only decreases the quantity of LNG in stock, but also changes its quality. Specifically, the composition of the fuel changes owing to what is called preferential evaporation (or as weathering in the LNG industry), which is a process in which the most volatile components of the substance evaporate first. This has a negative effect on the meth-ane number of the fuel, which is a quality measure for LNG as a fuel that serves as an indicator of the fuel’s knock resistance. If an engine runs on LNG with a lower knock
Introduction 7 resistance than given in the engine specification, there is a risk of engine knock (i.e. spontaneous ignition). This not only reduces fuel efficiency, but can also damage the engine in extreme cases. As a consequence, manufacturers of LNG engines strongly recommend end-users to ensure that their trucks or ships be fueled with LNG that meets the quality specification of the engine.
Theoretically, the methane number of LNG can range from 0 to 100, where 0 is pure hydrogen and 100 is pure methane. The methane number decreases when the concentration of light components of the fuel (e.g. methane and nitrogen) decreases with respect to that of heavier hydrocarbons (e.g. ethane and propane). Depending on the region in which the LNG is produced, the composition (and thus the quality) of LNG varies. For example, LNG from Nigeria, which is one of the main sources of LNG in Europe, has a methane number of 71 and that of Norway is 76. Interestingly, the methane number of Bio-LNG is higher than that of any fossil LNG in the market. In addition to the regional differences, the quality of LNG changes over time as it is moved through or stored in the LNG distribution network. A typical engine running on LNG requires a minimum methane number of 75 or 80, which is higher than the default of the LNG produced in several regions. Accordingly, alternatives to manage the quality of the LNG need to be considered in the small-scale distribution network of LNG.
One mechanism to ensure that LNG is sold at sufficient quality is to reduce the rate of quality deterioration by means of re-liquefaction of the boil-off gas. In large-scale import terminals, the boil-off gas generated can be used as a power source or can be re-liquefied. In the latter case, the boil-off gas is re-liquefied using specialized refrigerating equipment. Currently, this option is often economically or operation-ally prohibited for small-scale facilities (Hu et al., 2016). Even if it were, this alternat-ive would not provide any solution when the LNG imported is, by default, below the quality requirements of LNG-fueled trucks and ships. Another alternative to ensure sufficient quality is by mixing different loads of LNG. A load of low-quality LNG could be mixed with a load of high-quality LNG (e.g. Bio-LNG) in order to increase its quality. Such mixing of LNG can be implemented in a small-scale distribution network to upgrade the quality of the fuel when needed. Given certain conditions, mixing different loads of LNG is technically feasible.
Managers of satellite and/or refueling facilities of an LNG distribution network aim to ensure that the LNG offered to their customers meets the quality
specifica-tions of the market. Hence, inventory-related decisions in those facilities must take into account the quality of the fuel and the process by which it deteriorates. To deal with these issues, managers can exploit the mixability property of LNG to manage the quality of the fuel and to mitigate the effects of deterioration. This characteristic is particularly interesting considering that in some LNG markets, more than one source of LNG is available (e.g. fossil LNG and Bio-LNG), which allows fuels of dif-ferent quality to be offered. Consideration of these aspects in inventory management problems leads to new challenging decision models that have not been previously studied in the literature on inventory control problems. In this thesis, we developed three projects in which we study inventory systems inspired by the problems arising in an LNG distribution network.
In Chapter 3, we conduct an analytical study where we investigate optimal in-ventory control decisions considering the properties of LNG (namely quality deteri-oration and mixability). To that end, we study an inventory system that faces con-stant demand for the commodity. Replenishment orders can be placed with multiple suppliers, each of which can offer the product with different price and quality. In our analysis, we make a partial characterization of the optimal policy by proving a few structural properties of the optimal solutions. These properties are not only useful to design solution approaches to the problem, but also provide intuition on the effect of the characteristics of LNG on inventory control decision-making.
In Chapter 4, we address the lot-sizing problem faced by an LNG storage facil-ity in a region where multiple suppliers are available, each of which may offer the fuel with different quality and price. The problem entails finding a minimum-cost replenishment plan satisfying demands over a finite planning horizon, while meet-ing a minimum quality level. Initially, we formulate the problem as a mixed-integer non-linear program and approximate it with a mixed-integer linear program. These models are appealing, since they can be directly fed into commercial solvers, but they are computationally expensive. Then, we focus on classes of solutions that can be obtained in polynomial time and develop two heuristics with varying levels of com-plexity. We illustrate numerically that one of our heuristics provides high-quality solutions in short computational times.
In Chapter 5, we study an inventory control problem for LNG storage and re-fueling facilities in regions where only one supplier is available. In this problem, we seek to maximize the profit of an inventory facility that faces stochastic demand. We
Introduction 9 model and solve the problem by means of a Markov decision process (MDP) and study the structural characteristics of the optimal policy. The insights obtained in the analysis of the optimal policy are translated into a simple, yet effective, invent-ory control policy in which decisions (i.e. replenishment and/or removal) are driven by both the quality and the quantity of the inventories. We assess the performance of our policy by means of a numerical study and show that it performs close to optimal under a variety of system configurations.
1.3
Journal publications
Chapters 2, 3, 4 and 5 are based on the following published or working papers.
Chapter 2: Lopez Alvarez, J.A., P. Buijs, R. Deluster , L.C. Coelho, E. Ursavas. 2019. Strategic and operational decision-making in expanding supply chains for LNG as a fuel. Omega advance online publication.
Chapter 3: Lopez Alvarez, J.A., O.A. Kilic, ,P. Buijs, I.F.A. Vis 2020. Multiple supplier inventory control problem for deteriorating and mixable substances. Working paper.
Chapter 4:Lopez Alvarez, J.A., O.A. Kilic, P. Buijs, I.F.A. Vis 2020. Dynamic lot sizing with multiple suppliers for deteriorating and mixable substances. Working paper.
Chapter 5: Lopez Alvarez, J.A., P. Buijs, O.A. Kilic, I.F.A. Vis 2020. An inventory con-trol policy for liquefied natural gas as a transportation fuel. Omega advance online publication.
Other publications:
Post, R.M., P. Buijs, M.A.J. uit het Broek, J.A. Lopez Alvarez, N.B. Szirbik , I.F.A. Vis 2018. A solution approach for deriving alternative fuel station infrastructure requirements. Flexible Services and Manufacturing Journal 30(3), 592-607.
Buijs, P., J.A. Lopez Alvarez, M. Veenstra, K.J. Roodbergen 2016. Improved collabor-ative transport planning at Dutch logistics service provider Fritom. Interfaces 46(2), 119-132.
