International Journal of Supply and Operations Management

IJSOM
November 2017, Volume 4, Issue 4, pp. 341-358
ISSN-Print: 2383-1359
ISSN-Online: 2383-2525
www.ijsom.com

A Review Study: A Conceptual Model of the Maritime Inventory-Routing Problem

Ali Nazemi a, Sanaz Sheikhtajian a,* and Majid Feshari a

a
Faculty of Economics, Kharazmi University, Tehran, Iran

Abstract

The main aim of this paper is to present a conceptual structure for the Maritime Inventory-Routing Problem. We have
looked at the matter from the supply chain angel and summed up the key comprising elements and major hypotheses in
the framework of a conceptual model. We have reviewed the related literature in detail in a classified manner, separated
various issues, and eventually in accordance with the identified gaps, presented the grounds for the development of the
model in the same particular framework. What we deal with in this article is in fact the zero level of the Maritime
Inventory-Routing Problem while for focusing on higher levels it is possible to consider greater details by providing
arithmetic models on more comprehensive navigation of naval goods in a more compact and sold manner. According to
this review, most researches are deterministic at the tactical level, on the basis of discreet time and in the arc-flow
framework, generally solved by exact or heuristic methods.

Keywords: Maritime inventory-rutting Problem; Conceptual model; Supply chain.

1. Introduction

Close relations between the transportation and inventory management field have led many researchers to do solid studies
on both topics simultaneously in the framework of Inventory Routing Problem (IPR). The issues include solidification
and harmonizing (integration and coordination) between the inventory management and vehicle routing, a major part of
which is composed of the Vendor Managed Inventory (VMI) (Agra, et al 2013b).

Among the studies in this field we can refer to articles by Hoff Et al (2010), Coelho Et al (2014) and SteadieSeifi Et al
(2014), in which the authors tried to survey the issue of inventory routing from the general and public points of the view.
They also evaluated the entire related studies in a comprehensive classification.

Among the various models of transportation, the maritime transportation plays a high profile role in commercial
transactions since based on a report in 2012 (UNCTAD, 2012), some 80 percent of the world trade in the year 2011 was
done by the shipping industry; this proves the significance of this field of transportation. Parallel with the expansion of
inventory routing in this sector, the discussions on maritime inventory routing problem (MIRP) were pursued. Studies
by Christensen, Fagorholt Et al (2004, 2013), and Papageorgio et al. (2014c) have comprehensively summed up the
procedure of past and contemporary related studies.

The studies in this field can be classified around various axes, the first of which is the time horizon that includes the
strategic, tactical, and operational levels; a major part of the studies is conducted in the tactical level, as is shown in
Figure 1.

Corresponding author email address: [email protected]

341
 Nazemi ,Sheikhtajian and Feshari

 The strategic level studies are concerned with long-term decision-

Strategic makings in the infrastructure field
 The tactical model mainly focuses on taking optimum advantage of the
existing infrastructures through choosing services, allocating capacities
to orders, planning for trips and voyages and proper timing of them.
Tactical
 The operational level, side by side with the objectives set in the tactical
field, is after surveying the issues in the framework of the actual
conditions and examining issues such as dynamic and contingency
Operational programming. This is a very complicated field and it is in need of precise
algorithms with high calculation speeds.

Figure 1. Levels of Programming in a Chronological Order (Sang and Ferman, 2013)

Keeping in mind the importance of the discussions on maritime routing and their strategic value, in this paper we have a
different look at other studies in order to deduct a conceptual model for the maritime inventory routing problem. Although
many studies have been conducted on the MIRP, a significant gap is observable and the necessity of constructing a
conceptual model encouraged us to propose a model in the present paper. The primary goal is to present, define and
structure the MIRP. By accomplishing this goal, two main contributions are made. First, the topic is reviewed and based
on the review a conceptual model is proposed, and second, based on the framework main directions for future research
are discussed to pave the way for it. To do so, we have taken account of an initial framework based on the supply chain
in the field of maritime transportation and have developed the framework and presented it as a proper conceptual model
in accordance with the structure of model-making in the articles (inclusive of initial hypotheses, variables, parameters,
etc.) as well as the prevailing approaches to problem resolving.

The remainder of this paper is organized as follows: In section two the maritime inventory routing problem is presented
in its general form; in section three the conceptual model for the questions is elaborated based on the framework of the
maritime supply chain; in section four the aftermath of achieving the objective and the ways for solving the problem are
dealt with, and in section five the grounds for the development of the model are described based on the conducted studies
and the proposals made by the authors of this article. Finally, section six includes a summary of the discussions and the
conclusion.

2. Maritime Inventory routing problem

Maritime inventory routing problem serves the purpose of simultaneously solving both issues of inventory management
and routing problem in the maritime transportation sector. The positions of both issues in the maritime supply chain is
depicted in Figure 2.

Storage & Transportaion & Unloainding &

Supply Consunption
Loading Maintenance Storage

Figure. 2. Supply chain in the offshore (naval) sector

In various studies, there are different ranges of issues with the maritime inventory routing problem; in some cases, the
whole chain from supply (provision) to consumption has been taken into consideration and in other studies, the focus
has merely been on the transportation and inventory management. Going on with the issue, brief descriptions of the fields
are presented as follows in order to get familiar with the generalities of the issue:
 Supply (provision): In the production sector, the manufacturer produces the goods based on a specific rate (fixed /
variable rate). It must be noted that a relatively wide range of productions have been taken into consideration here
in the field including ammonia, cement, oil and petroleum extracts, and petrochemical products. The products are
temporarily stored and then loaded in somewhere close to the port terminal by special ships or tankers.
 Storage and loading: The manufactured products here are stockpiled in due places and will be stored based on some
policies for goods control and storage limitations. A shipping transportation fleet is employed in order to send off
goods from the points of production to various destinations. The said ships are also being loaded according to the
limitations of their entry into any ports.

Int J Supply Oper Manage (IJSOM), Vol.4, No.4 342

 A Review Study: A Conceptual Model of the Maritime ...

 Maritime Transportation: The ships head off to the destination ports after being loaded. The main aim here is to
define the route and departure time for the ships. The fleet of ships could have been in the possession of the
manufacturers or been chartered according to the manufacturer’s policies.
 Maintenance and repairs: The ships are subject to a maintenance program in specific time intervals which is the
basis for spewing them out of the list of available ships.
 Unloading (discharging) and storage: Ships enter the discharging ports based on the limitations and restrictions on
accepting a specific ship type as well as their admission and discharging time. The products are being stored in the
port before being consumed.
 Consumption: Consumers receive the products through two ways of long-term contracts or spot which have been
modeled with different assumptions to deal with different issues.

The main objective of the problem here is to determine the policies for routing and inventory management along with
the optimization of the manufacture’s objective function which has already been defined in a way that it could maximize
the profits or minimize the costs.

The policies will be finally implemented in different ways based on the related existing general restrictions in the field
of transportation including restrictions on the capacities of the goods and fulfilling the contractual obligations of the
optimization model. Pursuing solutions for this question, we have then gone over the main points of the studies in the
field and deducted the conceptual model accordingly.

3. A review of the literature for developing a conceptual framework

To build a conceptual framework we use Meredith (1993)’s philosophical conceptualization approach, which is based
on reading the papers over and over again (Seuring and Muller, 2008). We start with reviewing the conducted studies
and using the maritime supply chain as a basis to categorize them. Table 1 gives a holistic view of the reviewed studies.
From supply to consumption we summarize and highlight the main characteristics of each paper with regard to 3 main
categories including (1) structure of modelling, (2) variables and parameters, and (3) uncertainties in each segment. Then,
referring to each study's number (in Table 1) we classify the studies in a figure.

The second step towards conceptualization is to propose a model and framework. As mentioned before, the building
blocks of the proposed MIRP conceptual model is the maritime supply chain (Fig. 2). The origin and validation source
of this model is Papageorgio et al (2014c)’s paper in which the researchers categorized the MIRP studies in a systematic
approach and developed a MIRP library.

3.1. Supply
3.1.1. Structure of Model Making
Model-making in the supply section, considering the types of products and the expanse of the question, involves
production, extraction, and provision. Accordingly, we can put forward the model-making in three fields of product
types, diversity of products1, and production rate:
 Product Type: Because of the dominance of marine transportation, the strategic and important products which
are carried by ships are naturally the focus of attention of the researchers in the field. Among them we can refer
to the liquidated natural gas (LNG), crude oil, cement, liquid products, ammoniac, etc. Some articles have
already resolved the problem in its general form, regardless of any certain type of product.
 Number of the Products: One of the characteristics of model-making is concentration on whether a single
product or a combination of products are in question. Most of the research focuses on single-product
consignments.
 Production Rate: The production rate is considered as an exogenous variable in most studies, but in articles in
which the approach is based on a combination of discussions on production planning, routing and inventory
management the production rate is regarded as a self-generating or endogenous variable in the model-making
process. (Persson et al, 2005, Gronhaug et al, 2010) Those studies in which the production rate is considered as
an input parameter in the question can be classified into two separate groups.
 Some researchers assumed a fixed rate of production while some others assumed variable production rates in
accordance with the timespan or the production port.

1 By diversity of products we mean whether there is a single product or multiple types.

Int J Supply Oper Manage (IJSOM), Vol.4, No.4 343

 Nazemi ,Sheikhtajian and Feshari

Meanwhile, in some studies the researchers paid more attention to those factors leading to changes in the
production rate and applied model-making prediction methods. As the classification of the articles is shown in
the following chart based on the above-mentioned method, obviously using the production rate as a fixed and
externally-generating (exogenous) parameter can be referred to as a serious restriction on model-making.
Since the production rate influences the general structure of the question - supposing that it is fixed - the validity
and the generalizing capacity of the results are also affected.
A specific approach in this field is to use the stable methods for model-making in order to add to the credibility
of the results.
Although merging the production planning issue with the routing and inventory management makes the question
more complicated, it leads to its greater flexibility as well.

The reason is that the producers can regulate their production rate based on their commitments in the contracts and their
goods stored in the warehouses in a way to minimize their expenditures in various phases. Fig. 3 classifies researches in
this area according to their assumptions about product number and production rate.

Fix-Rate L-bulk[31],fuel-oil[1]
Single_Prod
uct LBulk[3,32],LNG[6,7,18,21,22,23,24,29,30,38,43],Fuel
Variable-
Oil[4],Crude[36],VGO[16,19,27,49],Ammonia[11],Ge
Production Rate
neral[25]

Fix_Rate LNG[41]
Multiple-
Product
variable- L-Bulk[5,28,37,40],LNG[6,8,34,35],Oil
Rate refineries[33],Cement[13],General[12,25]

Figure 3. Distribution of the Researches in the Production Sector

3.1.2. Variables and Parameters

Variables and parameters in the production section are summarized in the following table. According to the researcher’s
viewpoint and objective, he or she can use all or some of the introduced variables and parameters in his or her model.

Table 2. Parameters and Variables of the Production Sector

Parameters Variable
 Available Time for Production  Production Rate
 Specifications and Number of Production Terminals
 The Lost Production Expenditure
 Production Costs
 Production Rate*
* The production rate, as mentioned before, is categorized as a parameter in some research works and as a variable in the problem-
solving and model-making process in some other.

3.1.3. Uncertainty in the Production Sector

The most important parameter in the field under the influence of uncertainty is production rate, which is of course rarely
dealt with in the related studies. Among those who have paid attention to this parameter are Halvorsen et al, 2013b) who
have presented a discrete probability distribution for daily production rate as the percent of planned rates.

3.2. Storage and Loading

3.2.1. Structure of Model-Making
After the production or extraction of the product (in case the product is oil, gas, or their derivatives and the like),
appropriate with the adopted storage policies and in accordance with the certain loading program the products are stored
and forwarded to the planned destinations. Model-making in the field includes the control policies on inventory and stock
type.

 Lot-Sizing: One strategy for responding to orders is summing them all up and responding to them altogether,
which is done in various forms. In order to strengthen and improve formula-making for this purpose, in some
studies about model-making the issue of defining the lot-size is considered. In this type of model-making it is
proved that using the variables that show when the production and also the consumption occur leads to the
strengthening and improvement of the models and formulations. (Hewitt et al, 2013)

Int J Supply Oper Manage (IJSOM), Vol.4, No.4 344

 A Review Study: A Conceptual Model of the Maritime ...

Table 1. The MIRP studies in one view (Current studies’ results)

No. Author(s) Product *plan level *Tim *Model Method
e
S T O D C A P
1 Agra, et al, 2016 FUEL OIL T D C A BC-local search heuristic
2 Agra, et al, 2013a L-BULK T D A BC
3 Agra, et al, 2013b FUEL OIL O D C A BSSDP
4 Al-Khayyal and Hwang, 2007 LBULK T C A C-HEURISTIC
5 Al-Haidous, et al, 2016 LNG T D ADP
6 Andersson, et al, 2015a LNG T P BC
7 Andersson, et al, 2015b LNG T D P BC
8 Anderson, et al, 2010 LNG T D P MODEL ONLY
9 Arijit, et al. 2017 T D A PSO-CP
10 Chengliang, et al. 2017 T A Heuristic method
11 Christiansen, 1999 AMMONIA T C A P BPC
12 Christiansen and Fagerholt, GENERAL T C A construction-heuristic
2009
13 Christiansen, et al, 2011 CEMENT T GENETIC
14 Drazen-Papvic, et al, 2011 GENERAL O D P HEURISTIC
15 Dung-Ying & Chang, 2018 GENERAL A Heuristic method
16 Engineer, et al, 2012 VGO T D P BPC
17 Ethan, et al. (2018) liquid helium D p Integer programming
18 Fodstad, et al, 2010 LNG T D DEFALT SOLVER
19 Furman, et al, 2011 VGO T D A DEFALT SOLVER
20 Giami, et al, 2016 LNG O D A P DEFALT SOLVER
21 Goel, et al, 2015 LNG S D A MIP BASED LS
22 Goel, et al, 2012 LNG T D A C&I- HEURISTIC
23 Gronhaug, et al, 2010 LNG T D P BPC
24 Halvorsen, et al, 2013 LNG T D A DECOMPOSITION
25 Hemmati, et al, 2016 GENERAL T C A HYBRID-HEURISTIC
26 Hemmati, et al, 2015 GENERAL T C A 2 PHASE HEURISTIC
27 Hewitt, et al, 2013 VGO T D A BPGS
28 Li, et al, 2010 L-BULK O C A DEFALT SOLVER
29 Mutlu, et al, 2015 LNG T D P VRH_HEURISTIC
30 Nikhalat-Jahromi, et al, 2016 LNG O C P MILP
31 Papageorgio, et al, 2014a L-BULK S D A APPROXIMATE
32 Papageorgio, et al, 2014b L-BULK T D A BENDERS LIKE
33 Person and Gothe-Ludgren, Oil Refinery T D A P CG-HEURISTIC
2005
34 Rakke, et al, 2014 LNG T D A BPC
35 Rakke, et al, 2011 LNG T D P ROLLING HORIZEN
36 Rocha, et al, 2013 CRUDE T D A BB
37 Ronen, et al, 2002 L-BULK T D A GRASP
38 Shao, Furman, et al, 2015 LNG S D A LAGRANGIAN
HEURISTIC
39 Shen, et al, 2011 CRUDE S D P GENETC
40 Siswanto, et al, 2011 L-BULK O C A CIH
41 Stalhane, et al, 2012 LNG T D P MILP BASED LS
42 Song and Furman, 2013 VGO T D A FIX-RELAX
43 Uggen, Fodstad , et al, 2011 LNG T D P FIX-RELAX
44 Yonneng , et al, 2015 GENERAL T D C P MODEL ONLY
 Inventory Type: The stored products in the warehouses can be divided into two groups based on whether the
products in question are deteriorated while they are in the storage or not. It is interesting to note that this issue
is involved in the case of LNG product as some studies on this product defined the Boil-Off-Rate, which is
considered both fixed and variable and is included in the model-making.

Int J Supply Oper Manage (IJSOM), Vol.4, No.4 345

 Nazemi ,Sheikhtajian and Feshari

Lot-Sizing [16]

Deteriorating
[5],[6],[7],[8],[18],[20],[21],[22],[23],[29],[30],[35],
Not Lot-Sizing
Inventory [38],[41]

Lot-Sizing [4],[31],[32]
Non-
Deterirating
[1],[3],[5],[11],[12],[13],[19],[24],[25],[26],[27],
Not Lot Sizing
[28],[33],[34],[36],[37],[39],[40],[44],[45],[46]

Figure 4. Distribution of the Researches in the Storage and Loading sector

3.2.2. Variables and Parameters

The variables and parameters of the storage and loading procedure are categorized in Table 2. Some of the factors are
involved in the storing process and the others in the loading procedure.

3.2.3. Uncertainty Factor in Storage and Loading

One of the uncertain factors in the storage sector is the amount of goods in the case of deteriorating products. By
deteriorating products, we mean goods like the LNG whose quantity is decreased over time due to various factors such
as evaporation. Under such conditions the decreasing level of such products is an uncertain parameter, but it is generally
given a fixed rate in model-making in order to simplify the calculations.

Among the other factors of uncertainty in the field of loading and ports which is mainly dealt with at the levels of tactical
and operational studies is the ports’ conditions and their availability, which are subject to change during protests,
technical difficulties, etc.

Table 3. Parameters and Variables in the Storage and Loading Process

Parameters Variables
 Inventory initial or final level  Replenishment level
 Storage Capacity (Min & Max)  Safety stock level
 Storage costs  Spot cargo number
 Stock out costs
 Value of Remaining products
 Specifications of ports (number of berths,
entry limitations)
 Ports entry costs
 Loading cost
 Demurrage
 Waiting cost

3.3. Transportations, Maintenance and Repairs

3.3.1. Structure of Model-Making
The most important part of the problem is related to the transportations field, thus compared with other fields, model-
making in this field as well as in repairs and maintenance has more details including defining the type of shipping, fleet
composition, the structure of transportation, and discussions on repairs and maintenance.

 Type of Shipping: There are three major types of operations for commercial ships: Liner, tramp, and industrial
(Lawrence, 1972). In the first type, the ships initially move in a line in a fixed route based on a pre-defined
program similar to the city bus fleets. In the second type, i.e. tramp, the ships are in pursuit of available goods
for shipment, like moving taxies seeking passengers in their way, whose optimum objective is gaining maximum
benefits at any given moment, keeping track of their initial positions, the place for unloading and the goods’
conditions. In the third type, that is, the industrial operations, generally the owner of the operations is the owner
of the carried goods and he or she is certainly interested in controlling the ships used in carrying his or her
goods. The objective in this type is minimizing the expense. Among the clients of industrial shipment, we can
refer to the vertically solid companies such as the oil and chemical product companies (Christensen et al, 2013).
A major part of the studies in the field are a combination of the 2nd and 3rd types.

Int J Supply Oper Manage (IJSOM), Vol.4, No.4 346

 A Review Study: A Conceptual Model of the Maritime ...

 Fleet Composition: Depending on whether the shipping fleet is homogenous or heterogeneous, this type of
distinction is made. A heterogeneous fleet is one in which ships’ specifications including their volumes, speeds,
carrying costs, etc. differ while in a homogenous fleet all specifications of entire ships are exactly the same. A
major part of the studies is done and designed on heterogeneous fleets.
 Transportation Structure: In accordance with the problem type in forwarding from one port of origin to one
destination (1-1), or from one port of origin to a number of destinations (1-m), and eventually from various
ports of origin to several destination ports (m-m), different types of model-makings can occur.
 Transportation Type: Two major types in the field are the successive and not-successive move. The successive
move or split delivery in sea transportation means a ship being permitted to move from one port to several ports
in a one-trip-program. The not-successive move or ship means ships are not allowed to request visiting
numerous ports in one voyage before returning to their port of origin (Mutlu et al, 2015).
One of the facts considered in many maritime routing problems is that split delivery is not

Maintenance
Successive
No-Maintanance
1-1
Maintenance
Not
Successive
No-Maintanance [42]

Maintenance [35]
Transportatio Successive
n& 1-m No-Maintanance
Maintenance
Not
Maintanance [5],[6],[8],[18],[34],[41]
Successive

Maintenance [29],[32]
Successive
No-Maintanance [24],[40]
m-m
Maintenance [21],[30],[38]
Not
Successive
No-Maintanance [1],[3],[4],[13],[20],[22],[23],[25],[27],[38],[31],[33],
[36],[37],[39],[44]

Figure 5. Researches’ Transportations, Maintenance and Repairs section

3.3.2. Variables and Parameters

The variables and parameters related to ships’ transportations, specifications of the fleets, and planning for maintenance
and repairs are as follows:

Table 4. Parameters and Variables in Ships’ Transportations

Parameters Variables
 The number of ships classified according to  Tracing the ships’ navigation route (at what time,
ownership differences and types of contracts through which route and, towards which destination
allocating them to this purpose is each ship sailing?)
 Ships’ specifications including capacities, speeds,  Number of ships
and list of ports in which they can take side  The status of ships, from availability point of the
 Available and unavailable ships (in accordance with view
being out of order, or under repairs and in  The number of full,or empty storages and containers
maintenance procedure) of each ship
 Voyage costs and length of time

Int J Supply Oper Manage (IJSOM), Vol.4, No.4 347

 Nazemi ,Sheikhtajian and Feshari

3.3.3. Uncertainty in Transportation with Ships

The transportation sector, as the main ring in this chain, is affected by many uncertainty factors. One of the major
uncertainty factors is weather conditions. This factor leads to changes in the voyage length as well as fuel consumption
and is dealt with more than the other uncertainty factors in Cheng and Duran (2014), Shyshou et al, (2010), Halversen et
al, (2014, 2015), and Agra et al, (2013c, 2015, and 2016). Another factor in the field is ship breakdowns. Although
separate studies on this issue in maritime transport have been done, considering it in maritime inventory routing problems
is not very common.

3.4. Unloading and Storage

3.4.1. Structure of Model-Making
In the unloading process, like the loading port, the main issue is the method for controlling the quantity of products and
the related policies. Another issue is considering the buoy ports by the side of costal ports in some models.
 Storage in Unloading Ports: The inventory management in question is sometimes not merely restricted to the
provision port as in some cases the inventory management is considered in both the provision and demand
process and accordingly the studies are divided into two main groups.
In one group the inventory management is considered in the demand side while in the other it is not.
 Port Type: In the real world, in addition to the coastal ports, there are the buoy ports where goods are unloaded
before being delivered to the clients; the trip continues from there. Only in one article model-making based on
such details was observed. (Fodstad et al, 2010).
Distinctions between the articles based on the storage policies in destination ports are depicted in Fig. 6.

Demand side [1],[3],[4],[5],[7],[11],[12],[13],[16],[18],[19],[2

Inventory 1],[22],[23],[25],[26],[27],[29],[30],[31],[32],[3
management 3],[34],[35],[36],[37],[38],[39],[40],[41],[44]
Unloadin & Storage

Not Demand
side Inventory [6],[8],[6],[24],[28],[41]
Management

Figure 6. Distribution of the Researches in the Unloading and Storage Section

3.4.2. Variables and Parameters:

According to the specifications of the ports of loading and storage, the following factors are defined as bellow (Table 5).

Table 5. Parameters and Variables in Ports of the Loading and Storage Section
Parameters Variables
 berths’ specifications: Number of ports,  Unloading rate
capacities of each, loading restrictions  safety stock level
 Port visit charge  stock out
 Demurrage in port  Unloading time
 Storage capacity (minimum and maximum)
 Stock out cost
 Storage costs

3.4.3. Uncertainty in the Ports of Loading and Storage

In addition to conditions such as protests and technical difficulties, which we referred to in ports of loading section
(Agra et al, 2015), one of the uncertainty factors in the ports of unloading is due to not knowing exactly in which port
goods are to be unloaded since only geographical specifications of the destination are defined. The manufacturer acts
according to either his or her experience or the average of the existing ports in the region. The worst conditions are
considered and various scenarios are surveyed in accordance with the time of voyage and other conditions and then the
programming takes place.
Another difficulty in dealing with unloading ports is the holidays. Different methods can be applied to prevent the
reaching of ships to ports on holidays, such as imposing penalties for doing so. (Christensen, Fagerholt et al, 2002)

3.5. Consumption
3.5.1. Structure of Model-Making

Int J Supply Oper Manage (IJSOM), Vol.4, No.4 348

 A Review Study: A Conceptual Model of the Maritime ...

This part, which is the last ring in the chain of maritime supply chain, can be surveyed in three fields including
consumption rates, types of contracts (long term or spot), and the ways for delivering products to markets and their
distribution.
 Consumption Rare: Generally, the consumption rate is considered as the input to the problem. In problems in
which uncertain rates are considered it is possible to use prediction methods to predict the consumption rate
and insert it in the model. Of course, in studies in which the sphere of survey is the entire supply chain defining
the consumption rate is considered as an indigenous variable. (Grunhag et al, 2010)
 Delivery Types: Depending on the responsibilities of sellers and buyers in the process of transferring products
three common types can be surveyed:
o FOB: the responsibility of transferring is up to the buyer as of the time of loading the goods onto the
ship(s).
o CIF: the responsibility of provision, paying the transferring expense and insurance of the goods is up
to the seller.
o DES: the responsibility of providing the ships and delivery are up to the seller and the ownership of
the goods is transferred to the buyer as soon as they are unloaded and carried from the ship(s) to the
buyer port. In these contracts the insurance costs are paid by the seller.
Among the studies done in this field only one article has focused on this issue. (Nikhalat et al, 2016)
Fix
Consumption [1]
Rate
LTC contract
Varaible
[3],[4],[5],[7],[11],[16],[20],[22],[25],[27],[
Consumption
Consumption 36],[39],[40],[42],[44]
Rate
Fix
Consumption [31]
LTC & Spot Rate
Contract Varaible [12],[13],[18],[19],[21],[23],[24],[26],[29
Consumption ],[30],[32],[33],[34],[35],[40],[41],[43]
Rate
Figure 7. Distribution of the Researches in the Consumption Sector

3.5.2. Variables and Parameters

The variables and parameters related to the Consumption Section are as follows:

Table 6. Parameters and Variables in the Consumption Section

Parameters Variables
 Contract penalty due to delays, early delivery, or  Extent of violation of quantity of forwarded
failure to deliver the products in accordance with the goods
contract (over/under)  Total sales
 Spot sale costs
 Spot sale revenue (uncommitted, arbitrage, etc.)

3.5.3. Uncertainty in Consumption

Price and level of demand are two variables and uncertain factors in the consumption section. Among the studies carried
out in this field we can refer to those of Cheng and Duran (2004), Shyshou et al. (2010), and Alvarez et al. (2011).

3.6. The Conceptual Model for the maritime inventory routing problem
The main goal of this framework is to open possibilities of research rather than to categorize the related content of the
existing literature. Considering this and following the aforementioned steps drawn from the literature review for
conceptualization, we propose our model with the marine supply chain backbone. Based on the above review we identify
3 building blocks for the conceptual framework including modelling structure, variables plus parameters, and
uncertainties. Then, integrating these with then marine supply chain we construct a conceptual framework (Fig. 8).

The model has two main sections. In the first section (the first column in each part) we define inputs and outputs of the
model which are parameters and variables, respectively. The parameters are shown using straight lines and the variables
are shown with stars and dotted lines.

The second part is inclusive of various conditions which can be considered in the model-making of MIRP. For instance,
in the production sector it consists of 3 elements, namely product type, production rate, and product number. In the

Int J Supply Oper Manage (IJSOM), Vol.4, No.4 349

 Nazemi ,Sheikhtajian and Feshari

storage and loading section, the researcher should define the inventory type and inventory management strategies which
are the same decision points in the unloading and storage segment.

The main part is transportation and maintenance, consisting of 5 elements as shipping type, fleet composition,
transportation structure and type, as well as maintenance strategies; in the last ring which is consumption the modeler
should decide about consumption rate, contract type, and delivery approaches.
Maritime transport is significantly affected by the uncertainty factors, but this has been ignored in most research works.
In each group of the classified factors there can be uncertain factors included in the model, which are shown by triangles.

4. Objective functions and solving methods

4.1. Objective functions and constraints:
Supplier’s utility can be either modelled as maximizing profit or minimizing cost. Surveying the studies in this field
reveals that mainly the minimizing approach is applied in model-making, and in a limited number of cases using methods
for maximizing profit, the model-making is pursued. (Gronhaug et al. 2010, Fodstad et al, 2010, Hewitt et al, 2013, and
Anderson et al, 2015a).

The differences among the studies are either in the details of the expenses or revenue details in the models. Besides, the
parameters related to each are presented separately in the conceptual model. The surveyed constraints in the studies are
classified under five fields of constraints in production, storage and loading, transportations, repair and maintenance,
unloading and storage, and finally in the field of consumption and contracts, most of which are presented in the
framework of initial parameters in the model. The discrete time models present the planning horizon in a discrete manner
and assume that the events (including loading/unloading, etc.) occur in certain points of time while the continuous models
consider no restrictions on the occurrence of events. Of course, there are also some studies in which the discrete and
continuous time horizons were applied in hybrid models (Agra, Chistiansen, et al. (2016), (2013), Yonneng, et al.
(2015)). The arc-flow models include variables on decision making for the movement of the ships between the ports
whereas the path-flow models pursue the sequence of the ships’ frequenting to and from the ports. (Papageorgiou et al,
2014b) Although surveying the studies in the field indicates that the arc-flow models are more frequently used (Fig. 11),
Gronhaug et al (2010) claimed that “the advantage of path-flow models are that indicate and nonlinear constraints and
cost can easily be incorporated when generating the paths” (Papageorgiou et al, 2014). In two of those studies only the
model for the problem is presented and there is no solution for it (Andersson et al, 2010, Yongheng et al, 2015), but in
the other reviewed studies various types of methods are used for the maritime inventory routing problem.

The basis for classification has been gained from SteaddieSeifi et al. (2014)’s article. Accordingly, the methods applied
for solving the problem are divided into exact, appreciative, heuristic, meta-heuristic, and hybrid-heuristic types; in the
present paper another method which is commercial solvers is added to them. The articles whose models have used
commercial solvers like the CPLEX, GAMZ, etc. are included in the classification. It is noteworthy that such methods
for problem solving are appropriate for problems with smaller dimensions. Figure 8 shows the distribution method of
articles from the viewpoint of the mathematical model-making and solving methods.

Another aspect of model-making is uncertainty modelling, which is investigated in the articles. As mentioned before,
despite the significance of the uncertainty issue in the maritime inventory routing problem, the prevalent way for
model-making in this issue is in definite and static models, and there are a limited number of studies which have
discussed the uncertainty factor. In a general sum-up of the various methods used for handling uncertainty and
reaching a robust model, we can refer to the following five groups. ( Barnhart, C. & Laporte, G.2007).
 Simulation
 re-optimization of different scenarios or input parameters
 adding slack to the input parameters (e.g., service speed)
 deterministic models that incorporate penalties
 Stochastic optimization models

Int J Supply Oper Manage (IJSOM), Vol.4, No.4 350

 A Review Study: A Conceptual Model of the Maritime ...

Exact[3,4,24,27,32,34,36]
Approximite[31]
Heuristic[1,21,22,33,42]
Arc-Flow Metaheuristic[37]
Hybrid Heuristiv[38]
Defult Solver[19,20]

Discrete
Exact[8,16,23]
Approximit -
Heuristic[14,29,33,38,41,44]
Path-Flow Metaheuristic[39]
Hybrid Heuristc [43]
Defult Solver[20]

Mathematical Model

Exact[4,11]
Approximite -
Heuristic [1,5,12,40]
Arc-Flow Metaheuristi c -
Hybrid Heuristic[25,26]
Defult Solver[28]

Continiues
Exact[20]
Approximite -
Heuristic -
Path-Flow Metaheuristi c -
Hybrid Heuristic-
Defult Solver-

Figure 9. Distribution of the researches in the mathematical methods

5. Future Studies
A large number of studies have been done on the maritime inventory routing problem, but there are still lots of
unpaved paths for further research.
lot-sizing

m 0%
demandside
single-product

Fix
fix

S 3%
deteririating

5% 2%
nm 0% 83%
Ltc

11
varaiable

m 0%
variable

else

60% 36% ns 44%

nm 3%
notdemandside
lot-sizing
multiple-product

m 3%
nondeterirating

17%
varaiable Fix
fix

3% 8% s 3%
LTC &Spot

nm 0%
1m
variable

m 0%
else

32% 54% ns 50%

nm 20%
m 6%
s
nm 6%
mm
m 9%
ns
nm 50%

s:successive
ns:not successive 0%-25%
m:maintenance 25%-50%
nm:no maintenance 50%-100%

Figure 10. Distribution of the Studies on maritime inventory routing

Int J Supply Oper Manage (IJSOM), Vol.4, No.4 351

 Nazemi ,Sheikhtajian and Feshari

Production
Available
Rate
Production Time Production

Product type Product Number Production Rate

Operation
Production Rate
Cost

Ammonia

Multiple

Variable
Pulp,etc
Cement
Fuel-oil

Single
L-Bulk
Crude

VGO
LNG

Fix
Weighted
Lost Production
Production Terminals

Inventory Max Lng Left in Tanker

Inventory &Min level Vlue
Initial/
final level Storage & Loading
Storage Tank
Cost
Replenish Inventory Decision Inventory Typt
ment level Stock out Cost

Deteriarating
Deteriarating
Lot-sizing

Non-
Safty Port Visiit

else
Stock Cost
level

Loading Rate
Spot
Cargo LNG
Num. Waiting Cost Boil-off rate
Berth Spec(Num.,..)

Vessel Vessel
Routig(where, Num.(LTC,Spot,) Transportation & Maintanance
Vessel
Spec(Speed, Transportation
Fleet Transportation Maintanance
capacity) Shipping Composition Structure Type
Vessel
Num.(LTC
Homogenous
Hetrogenous

Susccessive
Successive
,Spot,)
Industrial
Tramp

Availabe/

Yes

No
M-M

Non-
1-M
Liner

1-1

Unavailabe
Vessel

Vessel
status(Ava
ilabe,..)
Transportation
Cost/Time Travel Fuel LNG
Vessel (load,..) Time Consumption Boil-off rate

Stock out cost

Port Visit
Unloaing
Cost Unloading & storage
Rate

weighting Cost
Saftey Unloading Port Port Type
stock
level Unloaing
Terminal Port
Management

Management

Bouy Port
No Inentory

Port(onShore,bouy)
Inentory

Stock out Berth

Spec(Num.,Capaciry )

Unloading LNG
time Inventory Max Storage Tank Cost
Boil-off rate
&Min level

Over/under
Cunsumption
Cost of violating
Delivery
inner time window
Penalty(exac
t/expected) Consumption
Contract Type Delivary
Rate
Variable

Over/under Spot sale costs

FOB

DES
Spot

CIF
LTC
Fix

Delivery )

Spot sale
Revenue(uncommited
Total sale
,arbitrage) Total
Spot-price
Demand

Figure 8. The Conceptual Model of the maritime inventory routing problem

Int J Supply Oper Manage (IJSOM), Vol.4, No.4 352

 A Review Study: A Conceptual Model of the Maritime ...

In the remaining part of this article we have examined them in the framework of implementing the supply chain. Based
on the distinctions made in the paper, the following sum-up about the percentage of articles on various fields is
presented and accordingly the fields on which less work has been done are categorized into three groups of less than
25%, between 25 and 50%, and above 50%. (Fig. 10)
5.1. Supply
As you notice from the supply section the majority of articles have considered the production rate as a variable factor
and have mostly focused on single-product goods. From the dimension of the required bases and involved concepts it is
possible to merge the following fields in the supply section:
 The sustainable production in case of the non-renewable resources.
 Production based on the environmental regulations while observing the limitations upon producing pollutants
and emissions.
 MIRP for a producer who plays in a carbon market.
 MIRP for a producer considering outsourcing policies.

5.2. Storage and Loading

In the storage and loading section, model-making is used regardless of strategies for lot-sizing as working on this issue
and modern inventory management models can pave the way for major developments in the field.
5.3. Transportations, Maintenance and Repairs
 In transportations, maintenance and repairs modelling many to many delivery without considering maintenance
and repairs is more prevalent. Besides, it is essential to pay more attention to repairs and maintenance as they
comprise an important part of the maritime transport.
Meanwhile, the other fields for development in this respect include:
 The green routing in maritime transportation
 Using fuzzy and other improved methods and analyzing techniques for decision making under unreliable
situations such as bad weather conditions, delays in time of voyage, etc.
 Using multi-function optimization models (inclusive of costs, social issues, pollution, risks involved in the
route, etc.)
 Merging the discussions on geographical navigations using software like the Arc-GIS in order to choose the
best routes for shipping during implementing the model under real-life conditions.

5.4. Unloading and storage

Just like supply ports, inventory management issue at the demand side is the focus of attention of model-makers. Similar
to the models applied in the loading side, using more modern methods of inventory management in this field is possible.

5.5. Consumption
The focus in the consumption phase is on the variable rate of consumption, and mainly long-term contracts are favored
by the model-makers, along with adding other models of contacts in baskets of contracts, in which development is a
matter of concern, too:
 Merging the pricing policies with discussions on inventory routing
 Maintaining balance in supply and demand using fuzzy methods for handling the uncertainties in pricing and
demand
 Considering the various conditions for delivery (FOB, CIF, etc.) in the model
 Merging the Games Theory with optimization models
 Using the nodal network models for predicting level of demand and prices in the market

6. Conclusion
Maritime transport plays a very decisive role in world trade, thus there is a lot of attention on it and a high number of
studies have been done on it. One of the most important characteristics of this field is the discussions on maritime
inventory routing, whose roots can be traced back to as early as 30 years ago in the studies of Bell et al, 1983. Since then,
this area of research has made a very satisfactory progress to the extent that many studies from various perspectives have
been done in the field.

In this article we have examined the maritime inventory routing problem from the viewpoint of supply chain, and by
elaborating the major and particular elements of the problem have presented a conceptual model for it. We have also
studied the structure of model-making and the various methods for solving the model under survey. The findings can be
categorized into two sections. First, we proposed a conceptual framework for modelling MIRP and then according to the
model, we classify current studies in order to identify research gaps in this field and highlight future research paths.

Int J Supply Oper Manage (IJSOM), Vol.4, No.4 353

 Nazemi ,Sheikhtajian and Feshari

According to the structure and statistical review of the studies (Fig. 10), there are many ways to develop researches with
a variety of strategies in each ring of marine supply chain from production to consummation, mainly in the third and the
last chains which are transportation, maintenance and consumption. Our review reveals that most of the studies were
conducted in the tactical level and they mostly modeled the arc-flow model which has been solved with heuristic and
exact methods.

To sum up, based on the results gained from most of the studied texts in the field of maritime inventory routing we can
say that most of research works are at the tactical levels, on the basis of discreet time, and in the framework of arc-flow
network models (Fig. 11); conducting strategic and operational level studies can be future developments in the field.
Meanwhile, as you noticed in the former section, the reason the majority of models are discrete is that the rate of
production and consumption is considered variable, which is in more in line with the reality.

Figure 11. Distribution of the Studies based on Model-Making and Time Horizon

Various methods have been adopted in order to solve this model, among which the heuristic method, the exact method,
and the commercial solvers have been the focus of attention of the researchers more than the others. In some of the
reviewed studies, the results of applying different methods are compared with each other. According to them, the speed
gained in the heuristic methods in comparison with the precise methods was more satisfactory, but the quality of their
answers was sometimes lower than that of the precise methods. Yet, using the precise models for solving complicated
problems is not always possible and this is one of the compelling reasons for using the heuristic methods. (Rakke et al,
2011, Goel et al, 2015, Agra et al, 2016).

Moreover, using the commercial solvers for complicated problems is not always possible, and consequently the gained
results are less competent than those of the other methods. (Goel et al, 2012; Rakke et al, 2014; Papageorgiou et al,
2014b)
Figure 12. Distribution of the Studies based on the Solving Method

model only defualt solver Exact Aproximation Hueristis Metaheuristic Hybrid Heuristic
Arc-Flow Path-Flow Arc & Path

Another point is the flexibility and generality of the models. The maritime inventory routing models have their own
specifications, and although their main structures have roots in the main body of inventory routing in consignments, their
particular features are still unique.

These issues are in fact the merged version of the two issues of inventory management, vehicle routing, and in some
cases production management. Thus, the relatively large number of parameters of the problems lead to complications in
these models. Of course, using the parameters in the form of couples of related parameters and observing their ratios of
effectiveness can decrease complications. (Mutlu et al, 2015). The last point is about the uncertainty discussions and the
robust nature of the models, to which, despite their significance in the field of maritime transport, less attention is paid
in the studies on the maritime inventory routing problems. They include some important uncertainty factors including
climatic changes (which lead to changing the voyage time and the extent of fuel consumption), difference in the quantity

Int J Supply Oper Manage (IJSOM), Vol.4, No.4 354

 A Review Study: A Conceptual Model of the Maritime ...

of the delivered products in deteriorating goods such as the LNG, the changing daily production rate proportionate to the
extent of the planning, the changes in the level of demand and also in the price, which need to be dealt with in the field.
Many researches have been done on the MIRP but there is still lot to develop. We categorized all future directions
according to our proposed model and distribution of the studies (Fig. 9) in section 5 which may be a good guide for
further studies in the field.

References
Arijit Dea, Sri Krishna Kumara, Angappa Gunasekaranb and Manoj Kumar Tiwaria (2017), Sustainable maritime
inventory routing problem with time window, Engineering Applications of Artificial Intelligence, Vol. 61, pp. 77–95

Agra, A., Christiansen, M., Hvattum, L. and Rodrigues, F. (2016). A MIP Based Local Search Heuristic for a Stochastic
Maritime Inventory Routing Problem. Springer International Publishing Switzerland, pp. 18-34

Agra, A., Christiansen, M., Delgado, A. and Hvattum, L. M. (2015). A maritime inventory routing problem with
stochastic sailing and port times. Computer and operations research. Vol. 61, pp. 18–30.

Agra, A., Andersson, H., Christiansen, M., and Wolsey, L. (2013a). A maritime inventory routing problem: Discrete
time formulations and valid inequalities. Networks, Vol. 62(4), pp. 297–314.

Agra, A., Christiansen, M., and Delgado, A. (2013b). Mixed integer formulations for a short sea fuel oil distribution
problem. Transportation Science, Vol. 47(1), pp. 108–124

Agra, A., Christiansen, M., Figueiredo, R., Hvattum, L.M., Poss, M. and Requejo, C. (2013c). The robust vehicle routing
problem with time windows. Computer and operations research. Vol. 40, pp. 856–866.

Al-Khayyal, F. and Hwang, S. (2007). Inventory constrained maritime routing and scheduling for multi-commodity
liquid bulk, part i: applications and model. European Journal of Operational Research, Vol. 176, pp. 106–130.

Al-Haidous, S., Msakni,M. and Haouari, M.(2016), Optimal planning of liquefied natural gas deliveries, Transportation
Research Part C, Vol. 69, pp. 79 –90

Alvarez, J. F., Tsilingiris, P., Engebrethsen, E., and Kakalis, N. M. P. (2011). Robust flee sizing and deployment for
industrial and independent bulk ocean shipping companies. INFOR, Vol. 49(2), pp. 93–107.

Anderson, H., Christiansen, M., and Fagerholt, K. (2010). Transportation planning and inventory management in the
LNG supply chain. In Bjørndal, E., Pardolos, P. M., and Ronnqvist, M., editors, Energy, Natural Resources and
Environmental Economics. Springer, Berlin.

Andersson, H., Christiansen, M., and Desaulniers, G. (2015a). A new decomposition algorithm for a liquefied natural
gas inventory routing problem. International Journal of Production Research, Accepted 29 Jan 2015, Published
online: 07 May 2015.

Andersson, H., Christiansen, M., and Desaulniers, G. and Rakke, J., (2015b). Creating annual delivery programs of
liquefied naturalgas, Springer Science+Business Media New York, Published online.

Barnhart, C., Laporte, G. (Eds.), Handbook in OR & MS, Vol. 14, 2007 Elsevier.

Chengliang Zhang, George Nemhauser, Joel Soko lexible (2017). Flexible Solutions to Maritime Inventory Routing
Problems with Delivery Time Windows, Computers and Operations Research, Vol. 89, pp. 153-162.

Cheng, L. and Duran, M.A. (2004). Logistics for world-wide crude oil transportation using discrete event simulation and
optimal control. Comput. Chem. Eng. Vol. 28, pp. 897–911.

Christiansen, M. (1999). Decomposition of a combined inventory and time constrained ship routing problem.
Transportation Science, Vol. 33, pp. 3–26.

Christiansen, M. and Fagerholt, K. (2002). Robust ship scheduling with multiple time windows. Naval Res. Logistics,
Vol. 49, pp. 611–625

Christiansen, M. and Fagerholt, K. (2009). Maritime inventory routing problems. In Floudas, C. A. and Pardalos, P.,
editors, Encyclopedia of Optimization, pp. 1947–1955. Springer.

Int J Supply Oper Manage (IJSOM), Vol.4, No.4 355

 Nazemi ,Sheikhtajian and Feshari

Christiansen, M., Fagerholt, K., Flatberg, T., Haugen, Ø., Kloster, O., and Lund, E. H. (2011). Maritime inventory routing
with multiple products: A case study from the cement industry. European Journal of Operational Research, Vol. 208(1),
pp. 86–94.

Christiansen, M., Fagerholt, K., Nygreen, B., and Ronen, D. (2013). Ship routing and scheduling in the new millennium.
European Journal of Operational Research, Vol. 228(3), pp. 467–483.

Christiansen, M., Fagerholt, K., and Ronen, D. (2004). Ship routing and scheduling: Status and perspectives.
Transportation Science, Vol. 38, pp. 1–18.

Coelho, L. C., Cordeau, J.-F., and Laporte, G. (2013). Thirty years of inventory routing. Transportation Science. (In
press).

Dung-Ying Lin and Yu-Ting Chang, (2018). Ship routing and freight assignment problem for liner shippiApplication to
the Northern Sea Route planning problem, Transportation Research Part E, Vol. 110, pp. 47-70

Popović, D., Bjelić, N. and Radivojević, G., (2011), Simulation Approach to Analyse Deterministic IRP Solution of the
Stochastic Fuel Delivery Problem, Procedia Social and Behavioral Sciences, Vol. 20, pp. 273–282

Engineer, F. G., Furman, K. C., Nemhauser, G. L., Savelsbergh, M. W. P., and Song, J. (2012). A branch-price-and-cut
algorithm for single product maritime inventory routing. Operations Research, Vol. 60, pp. 106–122.

Ethan Malinowskia, and Mark H. Karwana, José M. Pintob, Lei Sunc, (2018). A mixed-integer programming strategy
for liquid helium global, Transportation Research Part E, Vol. 110, pp. 168–188.

Fodstad, M., Uggen, K. T., Rّmo, F., Lium, A., and Stremersch, G. (2010). LNGScheduler: A rich model for
coordinating vessel routing, inventories and trade in the liquefied natural gas supply chain. Journal of Energy Markets,
Vol. 3(4), pp. 31–64.

Furman, K. C., Song, J.-H., Kocis, G. R., McDonald, M. K., and Warrick, P. H. (2011). Feedstock routing in the
exxonmobil downstream sector. Interfaces, Vol. 41(2), pp. 149–163.

Ghiami, Y., Woensel, T., Christiansen, M., and Laporte, G.,(2015), A Combined Liquefied Natural Gas Routing and
Deteriorating Inventory Management Problem, Springer International Publishing Switzerland 2015, F. Corman et al.
(Eds.): ICCL 2015, LNCS.

Goel, V., Furman, K. C., Song, J.-H., and El-Bakry, A. S. (2012). Large neighborhood search for lng inventory routing.
Journal of Heuristics, Vol. 18(6), pp. 821– 848.

Goel, V., Slusky, M., van Hoeve, W.-J., Furman, K., and Shao, Y. (2015). Constraint programming for lng ship
scheduling and inventory management. European Journal of Operational Research, Vol. 241(3), pp. 662–673.

Grønhaug, R., Christiansen, M., Desaulniers, G., and Desrosiers, J. (2010). A branch-and-price method for a liquefied
natural gas inventory routing problem. Transportation Science, Vol. 44(3), pp. 400–415.

Halvorsen-Weare, E. E. and Fagerholt, K. (2013). Routing and scheduling in a liquefied natural gas shipping problem
with inventory and berth constraints. Annals of Operations Research, Vol. 203(1), pp. 167–186.

Halvorsen-Weare, E. E., Fagerholt, K.: Robust supply vessel planning. In: Pahl, J., Reiners, T., Vo‫ك‬, S. (eds.) INOC
2011. LNCS, Vol. 6701, pp. 559–573. Springer, Heidelberg (2011).

Hemmati, A., Hvattum, L., Christiansen, M. and Laporte, G., (2016), An iterative two-phase hybrid matheuristic for a
multi-product short sea inventory-routing problem, European Journal of Operational Research, Vol. 252, pp. 775–788.

Hemmati , A., Stålhane, M., Hvattum, L. and Andersson, H., (2015), An effective heuristic for solving a combined cargo
and inventory routing problem in tramp shipping, Computers & Operations Research, Vol. 64, pp. 274-282.

Hewitt, M., Nemhauser, G. L., Savelsbergh, M., and Song, J. (2013). A branch and-price guided search approach to
maritime inventory routing. Computers and Operations Research, Vol. 40, pp. 1410–1419.

Int J Supply Oper Manage (IJSOM), Vol.4, No.4 356

 A Review Study: A Conceptual Model of the Maritime ...

Hoff, A., Andersson, H., Christiansen, M., Hasle, G., and Lّkketangen, A. (2010).Industrial aspects and literature
survey: Combined inventory management and routing. Computers & Operations Research, Vol. 37(9), pp. 1515–1536.

Li, J., Karimi, I., & Srinivasan, R. (2010). Efficient bulk maritime logistics for the supply and delivery of multiple
chemicals. Computers & Chemical Engineering, Vol. 34(12), pp. 2118–2128

Meredith J. (1993). Theory building through conceptual methods. International Journal of Operations & Production
Management, Vol. 13(5), pp. 3–11.

Mutlu, F., Msakni, M., Yildiz, H., Sonmez, E., Pokhare, S,. A Comprehensive Annual Delivery Program for Upstream
Liquefied Natural Gas Supply Chain, Journal of Operational Research, Vol. 250, pp. 120- 130.

Nikhalat-Jahromi, N., Bell, G., Fontes, D., Cochrane, R. and Angeloudis, P,. (2016), Spot sale of uncommitted LNG
from Middle East: Japan or the UK?, Energy Policy, Vol. 96, pp. 717–725.

Papageorgiou, D. J., Cheon, M.-S., Nemhauser, G., and Sokol, J. (2014a). Approximate dynamic programming for a
class of long-horizon maritime inventory routing problems. Transportation Science. Vol. 49, pp. 850- 885.

Papageorgiou, D. J., Keha, A. B., Nemhauser, G. L., and Sokol, J. (2014b). Two-stage decomposition algorithms for
single product maritime inventory routing. INFORMS Journal on Computing, Vol. 26(4), pp. 825–847.

Papageorgiou, D. J., Nemhauser, G. L., Sokol, J., Cheon, M.-S., and Keha, A. B. (2014c). Mirplib–a library of maritime
inventory routing problem instances: Survey, core model, and benchmark results. European Journal of Operational
Research, Vol. 235(2), pp. 350–366.

Persson, J. A. and Gothe-Lundgren, M. (2005). Shipment planning at oil refineries using column generation and valid
inequalities. European Journal of Operational Research, Vol. 163, pp. 631–652.

Rakke, J. G., Andersson, H., Christiansen, M., and Desaulniers, G. (2014). A new formulation based on customer
delivery patterns for a maritime inventory routing problem. Transportation Science, Vol. 49(2), pp. 384–401.

Rakke, J. G., Stalhane, M., Moe, C. R., Christiansen, M., Andersson, H., Fagerholt, K., and Norstad, I. (2011). A rolling
horizon heuristic for creating a liquefied natural gas annual delivery program. Transportation Research Part C, Vol. 19,
pp. 896–911.

Rocha, R., Grossmann, I. E., and de Arag‫م‬o, M. V. P. (2013). Cascading knapsack inequalities: reformulation of a crude
oil distribution problem. Annals of Operations Research, Vol. 203(1), pp. 1–18.

Ronen, D. et al. (2002). Marine inventory routing: Shipments planning. Journal of the Operational Research Society,
Vol. 53(1), pp. 108–114.

Seuring, S. and Muller, M., (2008). From a literature review to a conceptual framework for sustainable supply chain
management, Journal of Cleaner Production, Vol. 16, pp. 1699–1710.

Shao, Y., Furman, K. C., Goel, V., and Hoda, S. (2015). A hybrid heuristic strategy for liquefied natural gas inventory
routing. Transportation Research Part C: Emerging Technologies, Vol. 53, pp. 151–171.

Shen, Q., Chu, F., and Chen, H. (2011). A Lagrangian relaxation approach for a multimode inventory routing problem
with transshipment in crude oil transportation. Computers & Chemical Engineering, Vol. 35(10), pp. 2113–2123.

Shyshou, A., Gribkovskaia, I., & Barcelّ, J. (2010). A simulation study of the fleetsizing problem arising in offshore
anchor handling operations. European Journal of Operational Research, Vol. 203(1), pp. 230–240.

Siswanto, N., Essam, D., and Sarker, R. (2011). Multi-heuristics based genetic algorithm for solving maritime inventory
routing problem. In Industrial Engineering and Engineering Management (IEEM), 2011 IEEE International Conference
on, pp. 116–120. IEEE.

St¨alhane, M., Rakke, J. G., Moe, C. R., Andersson, H., Christiansen, M., and Fagerholt, K. (2012). A construction and
improvement heuristic for an lng inventory routing problem. Computers & Industrial Engineering, Vol. 62, pp. 245–
255.

Int J Supply Oper Manage (IJSOM), Vol.4, No.4 357

 Nazemi ,Sheikhtajian and Feshari

Song, J. and Furman, K. C. (2013). A maritime inventory routing problem: Practical approach. Computers and
Operations Research, Vol. 40, pp. 657–665.

SteadieSeifi, M.,Dellaert,N.P., Nuijten,W., Woensel, T. and Raoufi, R.,(2014), Multimodal freight transportation
planning: A literature review, European Journal of Operational Research, Vol. 233, pp. 1–15.

Uggen, K., Fodstad, M., and Nørstebø, V. (2013). Using and extending fix-and-relax to solve maritime inventory routing
problems. TOP, Vol. 21(2), pp. 355–377.

Yongheng, J. and Grossmann, I. E. (2015). Alternative mixed-integer linear programming models of a maritime
inventory routing problem. Computers & Chemical Engineering, Vol. 77, pp. 147–161.

European Commission. European transport policy for 2010: time to decide. White Paper. Luxembourg: Office for
Official Publications of the European Communities; 2001.
https://ec.europa.eu/transport/themes/strategies/2001_white_paper_en.

UNCTAD (2012). Review of maritime transport. United Nations, New York and Geneva.

Int J Supply Oper Manage (IJSOM), Vol.4, No.4 358