You are on page 1of 11

Computers and Chemical Engineering 34 (2010) 2118–2128

Contents lists available at ScienceDirect

Computers and Chemical Engineering


journal homepage: www.elsevier.com/locate/compchemeng

Efficient bulk maritime logistics for the supply and


delivery of multiple chemicals
Jie Li a , I.A. Karimi a,∗ , Rajagopalan Srinivasan a,b
a
Department of Chemical and Biomolecular Engineering, National University of Singapore 4 Engineering Drive 4, Singapore 117576
b
Process Sciences and Modeling, Institute of Chemical and Engineering Science, 1 Pesek Road, Jurong Island, Singapore 627833

a r t i c l e i n f o a b s t r a c t

Article history: Many multinational chemical companies (MNCs) manage the inventories of several raw materials at
Received 14 December 2009 their worldwide sites. Maritime transportation plays a key role in this chemical logistics. In this paper,
Received in revised form 20 July 2010 we address an inventory service problem in which a chemical MNC uses a fleet of multi-parcel ships
Accepted 26 July 2010
with dedicated compartments to move multiple chemicals continually among its internal and external
Available online 5 August 2010
production and consumption sites. The objective is to ensure continuity of operation at all sites by main-
taining adequate inventory levels of all raw materials. We develop a novel multi-grid continuous-time
Keywords:
mixed-integer linear programming (MILP) formulation based (Susarla, Li, & Karimi, 2010) for this chem-
Routing
Scheduling
ical logistics problem. Our model allows limited jetties at each site, non-zero transfer times, variable
Mixed-integer linear programming load/unload quantities, transfer task sequencing, etc. In contrast to the literature, it needs no separate
Transportation estimates for arrivals at each site. Several examples are solved to illustrate the efficiency of our proposed
Chemical logistics formulation.
Supply chains © 2010 Elsevier Ltd. All rights reserved.
Inventory management

1. Introduction Therefore, operational efficiency of maritime transportation can


reduce final material costs significantly. While transportation rout-
Logistics is the glue that binds global chemical supply chains. ing and scheduling has received much attention in the literature,
Maritime bulk transport is the workhorse of chemical logistics. relatively few papers have addressed ship routing and schedul-
Most MNCs (Multi-National Corporations) have plants distributed ing (Laporate & Osman, 1995). Christiansen, Fagerholt and Ronen
all over the world. A material of one plant may be a raw material for (2004) presented a detailed review on ship routing and schedul-
another that may be half-way around the world. Chemical mate- ing. Al-Khayyal and Hwang (2007) classified maritime ship routing
rialion is often far from demand locations or raw material supply and scheduling problems into cargo routing and inventory routing.
centers. For instance, the flurry of expansions in South East Asia, Because the aims of these two classes are to serve the cargos and
Asia-Pacific, Middle East, etc. related to manufacturing and logis- inventories respectively, we call them cargo service and inventory
tics activities is far from the major demand centers of US, Europe, service problems.
Japan, etc. The fast-expanding exporters (oil, gas, and chemicals) in The work on cargo service (Bausch, Brown, & Ronen, 1998;
the Caspian Sea, Africa, S America, and Arabian Gulf are also far from Brown, Graves, & Ronen, 1987; Fagerholt & Christiansen, 2000a;
their customers in East and West. For these reasons, and the pres- Fagerholt & Christiansen, 2000b; Fisher & Rosenwein, 1989; Jetlund
ence of centralized materialion in huge clusters such as Houston, & Karimi, 2004; Papadakis & Perakis, 1989; Ronen, 1986) involves
Rotterdam, Singapore, Jubail, etc. make the global maritime trans- ship routing and scheduling for a set of well-defined cargos of
port of chemicals inevitable and crucial to the success of chemical known quantities, loading/unloading ports, and time windows. The
enterprises and supply chains. Many major chemical companies aim is to either maximize profit or customer service. For instance,
own/charter multi-parcel ships or outsource to 3PLs/4PLs. Shipping Fagerholt and Christiansen (2000a, 2000b) addressed multi-ship
companies use multi-parcel chemical tankers to transport multi- pickup and delivery of dry bulk cargos with specified time win-
ple chemicals. Approximately 80% of all maritime ton-miles is via dows and pickup and delivery ports. Since each ship has a flexible
bulk (US DOT, 1999), which offers the lowest cost per ton-miles. cargo hold that can be divided into several small holds, they allowed
each ship to deliver multiple materials at the same time. Jetlund and
Karimi (2004) focused on multi-ship pickup and delivery of mul-
∗ Corresponding author. Tel.: +65 6516 6359; fax: +65 6779 1936. tiple liquid bulk cargoes with specified quantities, time windows,
E-mail address: cheiak@nus.edu.sg (I.A. Karimi). and pickup/delivery ports.

0098-1354/$ – see front matter © 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.compchemeng.2010.07.031
J. Li et al. / Computers and Chemical Engineering 34 (2010) 2118–2128 2119

plier ensures that no customer runs out of stock at any time. Miller
Nomenclature (1987) addressed a maritime routing problem in which a supplier
ships multiple chemicals from one port to multiple destinations to
Sets
maintain certain inventory levels. Other applications include fuel-
INT sites that belong to MNC
oil distribution (Dror & Ball, 1987), garbage collection (Russel & Igo,
Mi materials that site i can produce or consume
1979), and soft drink distribution (Golden & Wasil, 1987).
Recently, Christiansen (1999) addressed an inventory service
Subscripts
problem without a central supplier, which is slightly different from
i site
that defined by Dror and Ball (1987). A heterogeneous fleet of
j site
ships transports a single material among production and consump-
m material
tion ports to ensure adequate inventory level at each consumption
k slot
site. They assumed a single jetty at each port and constant load-
Superscripts ing/unloading and production/consumption rates, and allowed
U upper limit variable loading/unloading quantities. However, their model needs
L lower limit a priori estimates for possible visit times at each port.
Al-Khayyal and Hwang (2007) extended the work of
Parameters Christiansen (1999) to address multiple materials, dedicated
URvim unloading rate of material m at site i by ship v ship-compartments, and unlimited jetties at each port. They
LRvim loading rate of material m at site i by ship v also allowed sequence-independent times for material load-
Tv0 time at which ship v begins material transfer tasks ing/unloading transitions. Although they allowed multiple ships
at the first site to arrive at a port at the same time, they did not allow them to
TTvij voyage time to move from site i to site j of ship v load/unload simultaneously. Like Christiansen (1999), their model
STi setup time for transfer tasks at site i also needs estimates for possible visit times at each port. More
Rim production or consumption rate of material m at site importantly, however, as we show later, their model may fail to
i ensure adequate inventory levels at all times.
cvij total cost for ship v to move from site i to site j Li, Karimi and Srinivasan (2008) developed a mixed-integer
cv fixed cost of loading or unloading a material for ship linear programming formulation (MILP) for the inventory ser-
v vice problem addressed by Christiansen (1999) and Al-Khayyal
H planning horizon and Hwang (2007), however with several additional real-life fea-
Bi the number of jetties at site i tures. They allowed multiple materials, multiple jetties, variable
loading/unloading quantities, simultaneous loading/unloading,
Binary variables sequence of loading/unloading, etc. However, their formulation
xvik 1 if ship v is at or moving from site i during its slot k does not consider sequence-independent transition times. Fur-
yvmk 1 if ship v loads or unloads material m during its slot thermore, they used process slots (Liu & Karimi, 2007) in their
k slot-based formulation, which needs excessive computation time
to solve large-scale problems.
0–1 continuous variables In this paper, we address the problem of Li et al. (2008). How-
zvijk 1 if ship v is at or moving from site i in slot k and is ever, instead of using process-slots, we develop a novel multi-grid
at or moving from site j in slot (k + 1) model using unit-slots (Susarla, Li, & Karimi, 2010) with better com-
yevk 1 if ship v ends the transfer of a material at the end putational efficiency. We also incorporate sequence-independent
of its slot k setup times for transitions from one material to another dur-
ing loading/unloading. We solve several examples to illustrate
Continuous variables the efficiency of our proposed formulation. We also illustrate the
Tvk end time of slot k on ship v drawbacks of previous work using several examples. Thus, our
Tik end time of slot k on jetties at site i model is more general and more efficient than previous models
Timk end time of slot k on storage unit m at site i of Christiansen (1999), Al-Khayyal and Hwang (2007), and Li et al.
timk intermediate time for checking inventory in storage (2008).
m at site i during slot k
RLvimk duration for which ship v loads/unloads material m
at site i in slot k 2. Problem definition
qvimk amount of material m that ship v loads/unloads at
site i during slot k Consider a chemical MNC with plants or sites distributed glob-
Qvmk amount of material m onboard ship v at the end of ally. It needs M raw materials (m = 1, 2, . . ., M). A site may consume
slot k some of these materials. A site may also produce materials that
Simk stock level of material m at site i at the end of slot k are raw materials for other sites. A site may also need a raw mate-
bik the number of jetties in use just after Tik rial that no other site produces. In this case, the MNC sources it
TC total operating cost ($) from external suppliers. Thus, we consider two types of plants or
sites. An internal site belongs to the MNC, while an external site is
an independent supplier not owned by the MNC. Let I be the total
number of plants or sites (i = 1, 2, . . ., I). Let INT = {i|site i belongs to
The work on inventory service involves ship routing and MNC}, Mi = {m|site i consumes or produces material m}. Each site i
scheduling to maintain sufficient inventory levels of materials at has Mi = |Mi | storage tanks, one for each material in Mi .
all ports. The typical aim is to minimize cost and ensure continu- The MNC wishes to manage the inventories of all Mi materials at
ity of operations at various sites. Dror and Ball (1987) defined this each internal site i ∈ INT. To this end, it uses a heterogeneous fleet
as a distribution problem where each customer maintains a local of V ships (v = 1, 2, . . . , V ) with multiple dedicated compartments.
inventory and consumes a known amount each day. A central sup- These ships move continuously among I sites to transport various
2120 J. Li et al. / Computers and Chemical Engineering 34 (2010) 2118–2128

Fig. 1. Schematic of the supply and delivery of multiple materials for a chemical MNC.

materials and maintain the inventories of raw materials at each site. 3. Inventory levels at all internal sites are monitored and those at
A site i has Bi identical jetties. A ship may load or discharge a mate- all external sites are ignored.
rial via any jetty at a site. Thus, at most Bi ships may load/unload 4. Each ship can visit any site, internal or external.
materials at site i. Moreover, a ship can load or discharge only one 5. Each ship can carry any material.
material at a time through a jetty. In other words, it must load or 6. All ships end at some sites before the end of the planning horizon.
discharge materials one by one. 7. Ships load and unload materials at constant rates that depend
Fig. 1 shows a schematic of the above problem. At time zero, the on the material, site, and ship.
V ships are at some known positions with some materials of known 8. All Bi jetties at each site are identical.
amounts on board. The goal is to manage their movements over a
planning horizon of H so as to maintain adequate inventory levels
The objective is to minimize the total cost of supply and delivery
of all materials at all internal sites at all times. With this, we state
logistics for the MNC.
our problem as follows:
Given:
3. Motivating example
1. V ships, their initial positions, capacity limits, unload rates, and
materials on board initially and their amounts. This example involves two ships (v = 1 and 2, V = 2), two mate-
2. I sites, jetties at each site, load rates, and fixed costs of loading rials (m = 1 and 2, M = 2), three sites (i = 1, 2, and 3, I = 3), and two
and unloading at sites. jetties at each site (B1 = B2 = 2). Site 1 consumes material 1 and pro-
3. M materials, materials that each site may produce or consume, duces material 2. Sites 2 and 3 both consume material 2 and produce
initial inventories, inventory limits, and production or consump- material 1. The planning horizon is 2 days (H = 2). At time zero,
tion rate profiles. both ships are empty. Furthermore, ship 1 is at site 1 and ship 2
4. Site-to-site travel times and costs. is at site 3. Table 1 gives the complete data. The aim of this exam-
5. Setup times for switching from one material to another. ple is to show that (1) the model of Al-Khayyal and Hwang (2007)
6. Planning horizon H. is unable to guarantee sufficient inventory levels at all times and
(2) the detailed timings of loading and unloading operations are
essential.
Decide:
We note that the model of Al-Khayyal and Hwang (2007) does
not schedule the material transfer (loading and unloading) oper-
1. Detailed travel schedule (ports, their sequence, and times) for
ations in detail. Thus, it is not able to monitor inventory levels at
each ship.
the starts/ends of each material transfer. It checks inventory lev-
2. Materials and amounts that each ship loads/unloads at each visit
els only at the start and end of all transfer operations at a site.
at each site.
Thus, it cannot guarantee inventory levels during the transfer oper-
3. Loading or unloading sequence of materials.
ations at a site. This allows the possibility that the inventory may
4. Inventory profiles of materials at each internal site.
exceed the tank capacity or a stock-out may occur. To illustrate
these, we solve this example using the model of Al-Khayyal and
Assuming: Hwang (2007). We get the schedule in Fig. 2 as its optimal solu-
tion. Since their model ignores sequence of transfer operations, we
1. All parameters are deterministic. consider two scenarios (Fig. 3a and b) at site 1. In Fig. 3a, ship
2. Each site has a dedicated storage tank for a material that it han- 2 first unloads material 1 during [0.4, 0.5] day, and then loads
dles. material 2 during [0.5, 1] day. In Fig. 3b, it first loads material 2
J. Li et al. / Computers and Chemical Engineering 34 (2010) 2118–2128 2121

Table 1
Data for the Motivating Example.

Site/ship Suitable material Jetties Initial material (wu) Capacity (wu)

m1 m2 m1 m2

i1 m1–m2 2 10 15 12 21
i2 m1–m2 2 5 15 30 30
i3 m1–m2 2 15 20 15 30
v1 m1–m2 – 0 0 10 10
v2 m1–m2 – 0 0 10 30

Ship i→j Travel time (day) Travel cost (mu) i.m Loading/unloading rate (wu/day)

v1 i1 → i2 0.3 1 i1.m1 100


i1 → 3 0.3 1 i1.m2 50
i2 → i1 0.3 1 i2.m1 100
i2 → i3 0.3 1 i2.m2 100
i3 → i1 0.3 1 i3.m1 100
i3 → i2 0.3 1 i3.m1 100
v2 i1 → i2 0.3 1.5 i1.m1 100
i1 → i3 0.3 1.5 i1.m2 50
i2 → i1 0.3 1.5 i2.m1 100
i2 → i3 0.3 1.5 i2.m2 100
i3 → i1 0.3 1.5 i3.m1 100
i3 → i2 0.3 1.5 i3.m1 100

Scheduling horizon H = 2 days, ship v1 visits site i1 in slot 1 and TTv10 = 0, ship v2 visits site i3 in slot 1 and TTv20 = 0, site i1 consumes m1 at 10 wu/day, and produces m2 at
20 wu/day, site i2 produces m1 at 5 wu/day, and consumes m2 at 10 wu/day, site i3 produces m1 at 5 wu/day, and consumes m2 at 10 wu/day.

Fig. 2. Optimal schedule for the Motivating Example from Al-Khayyal and Hwang (2007).

during [0.4, 0.9] day, and then unloads material 1 during [0.9, 1.0] develop a new multi-grid formulation that would improve the solu-
day. tion efficiency.
The model of Al-Khayyal and Hwang checks inventories at t = 0.4
day and t = 1 day only. In Fig. 3a, these are 6 and 13 units for materi- 4. Mathematical formulation
als 1 and 2 respectively at t = 0.4 day, and 10 and 0 units respectively
at t = 1. While these are all within the desired limits, this does not We consider ships (v), jetties, and storage tanks as units (u). Since
hold at instances within [0.4, 1.0] day. For instance, the inventory of all jetties at each site i (1 ≤ i ≤ I) are identical, we treat them as
material 1 at t = 0.5 day is 15 units, which exceeds the tank capacity one single resource (i) with a capacity Bi . Since each storage unit is
of 12 units. Similarly, in Fig. 3b, the inventories of materials 1 and dedicated for a material, we treat each tank (m ∈ Mi at sites i = 1, 2
2 are 1 and (−2) units respectively at t = 0.9 day. . . ., I) as one distinct resource with stock limits (Sim L and Simu ). For
In contrast, the model of Li et al. (2008) schedules transfer oper- each unit u, we divide the planning horizon [0, H] into K (k = 1, 2,
ations in detail and does not face any such issues. It gives the correct . . ., K) ordered contiguous unit-slots (Fig. 6). We define Tu0 ≥ 0 as
optimal schedule (Fig. 4) and inventory profiles (Fig. 5a–b). the start of slot 1. Let Tuk (k = 1, . . ., K; Tuk ≤ H) denote the end time
Even though, the model of Li et al. (2008) works correctly, it of slot k on unit u. Slot k on unit u starts at Tu(k−1) and ends at Tuk .
requires excessive computation time because of its use of a single- We define the period before Tu0 as slot zero (k = 0). In contrast to Li
grid formulation (Susarla et al., 2010). Therefore, our goal is to et al. (2008), the slots are asynchronous, so the slot timings need
2122 J. Li et al. / Computers and Chemical Engineering 34 (2010) 2118–2128

Fig. 5. Inventory profiles of materials 1–2 for the Motivating Example from Li et al.
(2008).

Fig. 3. Inventory profiles of materials 1–2 at site 1 for the Motivating Example from
Al-Khayyal and Hwang (2007). Throughout this paper, each variable is defined for specific
ranges of its indices, and each constraint, unless otherwise indi-
cated, is written for all valid values of the indices of its constituent
not be same on all units (Susarla et al., 2010). Then, we have: variables.
Tvk ≥ Tv(k−1) 1≤k≤K (1)
4.1. Ship routing and scheduling
Tik ≥ Ti(k−1) 1 ≤ i ≤ I, 1 ≤ k ≤ K (2)
At time zero, a ship v may already be at a site or it may be at a
Timk ≥ Tim(k−1) i ∈ INT, m ∈ Mi , 1 ≤ k ≤ K (3) point in sea and heading towards a site. If the ship is at sea at time

Fig. 4. Optimal schedule for the Motivating Example from Li et al. (2008).
J. Li et al. / Computers and Chemical Engineering 34 (2010) 2118–2128 2123

(including zero and I + 1 sites), so


I+1
xvik = 1 1≤k≤K (4)
i=0

Once a ship reaches zero site, it cannot visit any other real site and
I + 1 site. Hence, zv0ik = 0 for 1 ≤ i ≤ I + 1 and 1 ≤ k < K. A ship always
first visits (I + 1) site. Thus, zvi(I+1)k = 0 for 1 ≤ i ≤ I and 1 ≤ k < K. Fur-
thermore, zvijk = xvik xvj(k+1) implies the following.


I+1
zvijk = xvik 0 ≤ i ≤ I + 1, 0 ≤ k < K (5a)
j=0

Fig. 6. Unit-slot design.



I+1
zvjik = xvi(k+1) 0 ≤ i ≤ I + 1, 0 ≤ k < K (5b)
j=0
zero, then we assume that it is heading towards a known destina-
tion site, where it will begin transfer tasks at Tv0 . If it is already at a While we treat xvik as binary variables, Eq. (5) enable us to treat zvijk
known site at time zero, then we take Tv0 = 0. In other words, slot as 0–1 continuous.
1 for each ship begins with the start of its transfer operations at
the first site, which we assume is known at time zero. While the
first arrival site is known for each ship, we do not assume anything 4.2. Material transfer operations
about the transfers that the ship will perform at that site. Similarly,
we assume nothing about where the ship will proceed from that To model the material transfer operations of a ship v, we define
first site. At each site, we allow a ship to load or unload materi- a binary variable (yvmk ) as follows,
als in several consecutive slots. A material transfer may occur over 
1 if ship v loads or unloads material m during
multiple slots. However, we do not allocate a separate or additional
yvmk = its slot k 1 ≤ v ≤ V, 1 ≤ m ≤ M, 1 ≤ k ≤ K
slot for the site-to-site movement of a ship. In other words, if a ship
0 otherwise
ends its transfer tasks in a slot k, then it moves to the next site in
the same slot. Fig. 7 illustrates the task log for a ship. Finally, when
If a ship is loading/unloading a material during slot k, then it must
a ship ends all its movements and transfers during the scheduling
be at or moving away from a site that can consume or produce that
horizon, then we say that it moves to a dummy site (i = 0), which
material. Similarly, if a ship is at a site, then it must load or unload
is at a zero distance from each real site (i = 1, 2, . . ., I). We also use
at least one of the materials that this site can consume or produce.
another dummy site (i = I + 1) to denote the initial position of a ship.
Lastly, a ship v must either load/unload a material or be at zero site
Thus, we now have I + 2 (i = 0, 1, 2, 3, . . ., I, I + 1) sites including I real
during each slot. Therefore, we write,
sites, zero and I + 1 sites. Once a ship reaches the zero site, then it
does not move from here. We assume that material transfer oper- 
xvik ≥ yvmk 1≤k≤K (6a)
ations on a ship begin at the start of a slot, but may end at any time
i:m ∈ Mi
within a slot.
Now, to track a ship’s position and movement from one site to 
another site with time, we define the following. yvmk ≥ xvik 1 ≤ i ≤ I, 1 ≤ k ≤ K (6b)
m ∈ Mi

1 if ship v is at or moving from site i during
xvik = its slot k 0 ≤ i ≤ I + 1, 1 ≤ k ≤ K 
M

0 otherwise xv0k + xv(I+1)k + yvmk = 1 1≤k≤K (6c)


m=1

 Note that Mi does not include the zero and I + 1 sites.


1 if ship v moves from site i to j at the end
Let RLvimk as the duration for which ship v loads/unloads mate-
zvijk = of its slot k 0 ≤ i ≤ I + 1, 0 ≤ j ≤ I + 1, 1 ≤ k < K
rial m at site i in slot k. If a ship v does not load/unload a material m
0 otherwise
during slot k, then these durations must be zero.
Note that zvijK is not defined, and zvijk = xvik xvj(k+1) . 
RLvimk ≤ H · yvmk 1≤k≤K (7a)
Based on the initial known position for each ship, we fix xvi0 .
During every slot, a ship v must be at or moving away from a site i:m ∈ Mi

Fig. 7. Schematic of slots and ship activities.


2124 J. Li et al. / Computers and Chemical Engineering 34 (2010) 2118–2128

Similarly, if a ship v is not at site i during slot k, then its loading or a resource, then the unit-slots in which this exchange occurs must
unloading durations must also be zero. have the same index on both the processing unit and the resource
 unit. In this case, ship is like a processing unit and jetties at each site
RLvimk ≤ H · xikv 1 ≤ i ≤ I, 1 ≤ k ≤ K (7b) are like a resource unit. Thus, whenever a ship v unloads/unloads a
m ∈ Mi material at a site during a slot, then the jetty resource at that site
must have a matching slot with the same index.
Let qvvimk be the amount of material m that ship v loads/unloads
Consider a ship v moving from a site j to site i during slot k. For
at site i during slot k. This is positive, if a ship unloads to the site,
the ship to begin its transfer tasks at Tvk , site i must assign one jetty
and negative, if it loads from the site. If the constant unloading and
before Tvk − STi . Since this jetty allocation occurs during slot k of
loading rates are URvim and LRvim respectively, then
ship v, we ask that it also occur during slot k of site i or at Tik. Since
−LRvim RLvimk ≤ qvimk ≤ URvim RLvimk m ∈ Mi , 1 ≤ k ≤ K (8) the jetty must be assigned to ship v before Tvk − STi , the following
must hold.
Before loading or unloading a material at a site, a ship may need ⎛ ⎞
some setup time. We assume for simplicity that this setup time 
I+1
(STi ) depends only on the site and not on ship or materials. Since a Tik ≤ Tvk − STi + H ⎝1 − zvjik ⎠ 0≤k<K (14a)
ship v may perform transfer operations in several consecutive slots, j=1,j =
/ i
we need to know the slot in which it completes them. To this end,
we define a 0–1 continuous variable yevk as follows, Now, consider site j. Before the ship moves away from site j
 during slot k, it must end all its transfer operations, and then
1 if ship v ends the transfer of a material at
release the jetty at site j. This release can occur only after Tv(k−1) +

yevk = the end of its slot k 1≤k≤K
m∈M
RLvjmk during slot k of ship v. Therefore, we demand that
0 otherwise j
the same occur during slot k of site j as well. If Tj(k−1) is the time at
A ship v cannot end a transfer in slot k, if it is at the zero site. which this jetty release occurs, then we must have,
⎛ ⎞
yevk + xv0k ≤ 1 1 ≤ k ≤ K (9)
 
I+1

If a ship loads or unloads a material in a slot k, but does not continue Tjk ≥ Tv(k−1) + RLvjmk −H ⎝1− zvjik ⎠ 1≤k<K (14b)
the same task in the next slot, then its run must end at slot k. m ∈ Mj i=0,i =
/ j

yevk ≥ yvmk − yvm(k+1) 1≤k<K (10a) Having aligned the jetty use by ships, we now do a resource balance
at each site i to ensure that no more than Bi jetties are in use during
yevK + xv0K = 1 (10b)
any slot. To this end, let bik (0 ≤ bik ≤ Bi ) denote the number of jetties
Similarly, if a ship loads or unloads a material in slots k and k + 1, in use just after Tik . A simple jetty balance over slots at site i gives
then it cannot end that task at the end of slot k. us,
 ⎛ ⎞
yevk ≤ 3 − yvmk − yvm(k+1) − zviik 1≤k<K (11) 
V 
I+1 
I+1

i:m ∈ Mi bik = bi(k−1) + ⎝ zvjik − zvijk ⎠ 1 ≤ i ≤ I, 1 ≤ k ≤ K


v=1 j=1,j =
/ i j=0,j =
/ i
If a ship is moving from a site to another in a slot k, then it must
also end all its material transfer tasks at the end of slot k. (15a)


I
yevk + zviik ≥ 1 1 ≤ k < K (12)

V 
I+1
i=0 bi(k−1) ≥ zvijk(k+1) 1 ≤ i ≤ I, 1 ≤ k ≤ K (15b)
From Fig. 7, when a ship ends all its transfer tasks at a site, then v=1 j=0,j =
/ i
it must move to another site, which may be the zero site. Let TTvij
We now use similar arguments to monitor material inventory in
(TTvi0 = 0, i = 0, 1, . . ., I) denote the time to move from site i to site
each storage tank.
j. After moving from site i, it will need some time (STi ) to set itself
up for transfer tasks. Therefore,
4.4. Inventory management

I 
Tvk − Tv(k−1) ≥ RLvimk Here a ship is a processing unit, storage tank is a resource unit,
i=1 m ∈ Mi and material in the tank is a resource. Now, consider a ship v that
is involved in a transfer operation of material m during slot k. Since

I 
I
this continuous-flow operation begins at the start of slot k on ship
+ (TTvij + STj )zvijk 1≤k≤K (13a)
v, we demand that it also begin at the start of slot k of tank m at site
i=1 j=1,j =
/ i
i.
If a ship v is loading or unloading at site I during slot k but not end at
Tim(k−1) ≤ Tv(k−1) + H[3 − xvik − yvmk − yev(k−1) ] i ∈ INT,
the end of slot k, then the duration for which ship v loads/unloads
must equal to the length of slot k on ship v. m ∈ Mi , 1 ≤ k ≤ K (16a)

I
RLvimk ≥ Tvk − Tv(k−1) − H · yevk 1≤k≤K (13b)
i=1 m ∈ Mi Tim(k−1) ≥ Tv(k−1) − H[3 − xvik − yvmk − yev(k−1) ] i ∈ INT,

m ∈ Mi , 1 ≤ k ≤ K (16b)
4.3. Jetty management
The above transfer operation may end before slot k of ship v ends.
For a formulation based on unit-slots, Susarla et al. (2010) Thus, we demand that it also end before slot k of tank m at site i
argued that if a processing unit and a resource unit are exchanging ends. Let Tim(k−1) ≤ timk ≤ Timk be the time at which this occurs. Then,
J. Li et al. / Computers and Chemical Engineering 34 (2010) 2118–2128 2125

Table 2
Data for Example 1.

Site/ship Suitable material Jetties Initial material (wu) Capacity (wu) Heel (wu) Production/consumption rate (wu/day)

i1 m1 1 20.4 80 1 2
i2 m1 1 60 100 1 1.25
i3 m1 1 6 25 1 2
i4 m1 1 – – – –
i5 m1 1 – – – –
v1 m1 – 0 15 0 –
v2 m1 – 0 24 0 –

Ship i→j Travel time (day) Travel cost (mu) i Loading/unloading rate (wu/day)

v1 i1 ↔ i2 6 8 i1 12
i1 ↔ i3 5.4 8 i2 12
i1 ↔ i4 3.95 5 i3 12
i1 ↔ i5 4.5 5 i4 12
i2 ↔ i3 2.2 8 i5 12
i2 → i4 7.5 8 – –
i2 → i5 6.9 8 – –
i3 ↔ i4 3.25 9 – –
i3 → i5 1.57 8 – –
i4 → i2 7.5 5 – –
i4 ↔ i5 2.6 5 – –
i5 → i2 6.9 5 – –
i5 → i3 1.57 5 – –
v2 i1 ↔ i2 6 3 i1 12
i1 → i3 5.4 9 i2 12
i1 ↔ i4 3.95 5 i3 12
i1 ↔ i5 4.5 5 i4 12
i2 → i3 2.2 4 i5 12
i2 → i4 7.5 6 – –
i2 → i5 6.9 7 – –
i3 → i1 2.5 9 – –
i3 → i2 2.2 9 – –
i3 → i4 3.25 9 – –
i3 → i5 1.75 9 – –
i4 → i2 7.5 5 – –
i4 → i3 3.25 5 – –
i4 → i5 2.6 5 – –
i5 → i2 6.9 5 – –
i5 → i3 1.75 10 – –
i5 → i4 8 – –

Ship v1 visits i5 in slot 1, and TTv10 = 0, Ship v2 visits i2 in slot 1 and TTv20 = 2.2 days, Scheduling horizon H = 16 days.

we demand that this time match the time at which ship v ends its which site i consumes or produces material m during the horizon.
transfer operation. Rim > 0 (Rim < 0), if site i produces (consumes) material m. Eqs. (16)
and (17) enable us to write the following mass balance for each
timk ≤ RLvimk +H(3−xvik − yvmk − yevk ) i ∈ INT, m ∈ Mi , 1 ≤ k ≤ K material tank.
(17a) Simk = Sim(k−1) + Rim [Timk − Tim(k−1) ]


V

timk ≥ RLvimk −H(3−xvik − yvmk − yevk ) i ∈ INT, m ∈ Mi , 1 ≤ k ≤ K + qvimk i ∈ INT, m ∈ Mi , 1 ≤ k ≤ K (19)


v=1
(17b)
Note that the MNC is not concerned about maintaining stocks at
Note that Eqs. (16) and (17) ensure that timk = Timk , whenever mul- external sites. Since Eq. (19) does not compute the stock level at
tiple ships are loading/unloading from/to the tank during a single intermediate points (timk ), we further need,
slot, and timk ≤ Timk , whenever only one ship is loading/unloading

V
from/to the tank. L
Sim ≤ Sim(k−1) +Rim [timk −Tim(k−1) ]+ U
qvimk ≤ Sim i ∈ INT, m ∈ Mi ,
For monitoring the inventory of material m onboard ship v, let
v=1
Qvmk (k = 0, 1, . . ., K; QvLm ≤ Qvmk ≤ QvUm ) be the amount of material
m onboard ship v at the end of slot k. Then, the mass balance for 1≤k≤K (20)
material m on the ship gives us, If required, one may want to ensure that the stock levels at the
 horizon end must also be within limits.
Qvmk = Qvm(k−1) − qvimk 1≤k≤K (18)
L U
i:m ∈ Mi Sim ≤ SimK + Rim (H − TimK ) ≤ Sim i ∈ INT, m ∈ Mi (21)

where Qvm0 is the known amount of material m on ship v at time 4.5. Objective function
zero.
For monitoring the inventory in each tank, we define Simk (Sim L ≤
This being a service problem, the objective is to minimize the
U ) as the stock level of material m at site i at the end of
Simk ≤ Sim total operating cost over the planning horizon. The key cost com-
slot k and Sim0 as the initial stock. Let Rim be the constant rate at ponents are transport (fuel and daily operation) costs, material
2126 J. Li et al. / Computers and Chemical Engineering 34 (2010) 2118–2128

Table 3
Model performance for the Motivating Example and Examples 1–3.

Example Model Discrete variables Continuous variables Constraints CPU time (s) MILP solution Relative gap (%)

Motivating Example Ours 40 490 728 0.48 7 0


Li et al. (2008) 120 831 1050 0.80 7 0
Al-Khayyal and Hwang (2007) 158 161 719 – N/A 0
1 Ours 44 667 609 0.17 32.5 0
Li et al. (2008) 106 967 905 3.25 32.5 0
Christiansen (1999) 312 147 1134 3.64 32.5 0
2 Ours 257 3791 3057 2351 22 0
Li et al. (2008) 597 4861 4301 86400 28 14.03
3 Ours 122 1575 1543 141 50.4 0
Christiansen (1999) 952 341 3317 1481 50.4 0

N/A, the obtained solution is infeasible.

transfer (site operations, duties, etc.) costs. Let cvij be the total cost the objective function does not change. For our case, we estimate
for ship v to move from site i to site j, and cv be the fixed cost of the number (K) of slots. We begin with a reasonably large number
loading or unloading a material for ship v. Then, the total cost (TC) and then check if the objective function changes, when we increase
is: K by 1.
K−1
We used CPLEX 12.1.0/GAMS 23.2.1 on a Dell workstation

V  
I+1 I+1
 
K
PWS690 (Intel® XeonTM CPU 3.00 GHZ, 16 GB memory) running
TC = cvij zvijk + cv yevk (22)
Windows XP for solving all the examples. Table 3 gives the model
v=1 i=0 j=0,j =
/ i k=1 k=1
performance for the examples.
Eqs. (1)–(22) complete our novel model for the inventory service
problem. While we assumed constant production/consumption 5.1. Motivating Example
rates for the sites during the horizon, a multi-period formulation
that allows piecewise constant production/consumption rates can We obtain the optimal solution (Fig. 8) of 7 mu (monitory units)
be derived from this single-period formulation in a straightfor- in 0.73 CPU s. Ship 1 loads 10 wu (weight units) of material 2 at
ward manner. The extension to the multi-period scenario should site 1 during [0, 0.2] day. Then, it moves to site 3, and unloads
be similar to that of Li et al. (2010). 10 wu of material 2 during [0.5, 0.6] day. Ship 2 loads 10 wu of
material 1 at site 3 during [0, 0.1] day. Then, it sails to site 1
5. Numerical evaluation and reaches there at 0.47 day. It loads 24 wu of material 2 dur-
ing [0.47, 0.95] day, and then unloads 10 wu of material 1 during
We solve four examples including the earlier Motivating Exam- [0.95, 1.05] day. For there, it moves to site 2 and unloads 24 wu
ple to evaluate our proposed model. These examples vary in the of material 2 during [1.35, 1.59] day. Fig. 9 shows the inventory
numbers of ships, sites, materials, jetties, etc. We give the data profiles of materials at each site. In Fig. 9a, the stock of material
(Table 2) for Example 1 only because of space considerations. The 1 at site 1 at 0.95 day is about 0.5 wu. Ship 2 begins replenish-
models of Christiansen (1999) and Al-Khayyal and Hwang (2007) ing that stock from 0.95 days. Since its unloading rate exceeds the
need an estimate of arrivals at each site separately. Thus, for solving site consumption rate, the stock increases from 0.5 to 9.5 wu at
their models, we first assume that the estimated number of arrivals 1.05 day. This stock is sufficient for site 1 until the horizon end
is the same for all sites. Then, we assume a small number for this (2 day), when its inventory is zero. By examining the stock pro-
arrival frequency, and then increase this number one at a time, until file for each material at each site, we conclude that the model

Fig. 8. Optimal schedule for Motivating Example from our model.


J. Li et al. / Computers and Chemical Engineering 34 (2010) 2118–2128 2127

Fig. 9. Inventory profiles of materials 1–2 for the Motivating Example from our model.

ensures that physical limits on the stock are respected at all consumption sites. The planning horizon is H = 16 days. We obtain
times. This is in contrast to the model of Al-Khayyal and Hwang the optimal solution of 32.5 mu within 0.17 CPU s, while the model
(2007). of Li et al. (2008) needs 3.25 CPU s. Since only one ship is involved in
this example, the model of Al-Khayyal and Hwang (2007) is reduced
5.2. Examples 1–3 to the model of Christiansen (1999). We obtain the optimal solution
of 32.5 within 3.64 CPU s from the model of Christiansen (1999).
Example 1 is from Li et al. (2008). It involves two ships, five sites Example 2 involves five ships, eight sites (five internal), and two
(three internal, two external), and one material. Among the five materials. The scheduling horizon is 80 days. Our model gives the
sites, sites 2, 4, and 5 are production sites, and sites 1 and 3 are optimal solution of 22 mu in 2351 CPU s. However, the model of Li
2128 J. Li et al. / Computers and Chemical Engineering 34 (2010) 2118–2128

et al. (2008) gets a suboptimal solution of 28 mu with a relative gap References


of 14.03% after 24 h of CPU time. Since the model of Christiansen
(1999) is for one material only and the model of Al-Khayyal and Al-Khayyal, F., & Hwang, S. J. (2007). Inventory constrained maritime routing and
scheduling for multi-commodity liquid bulk, part I: Applications and model.
Hwang (2007) may violate stock limits, when two or more materials European Journal of Operational Research, 176, 106–130.
are involved, we did not solve this example with their models. Bausch, D., Brown, G., & Ronen, D. (1998). Scheduling short term marine transport
Example 3 involves five ships, five sites, and one material with of bulk materials. Maritime Policy and Management, 25, 335–348.
Brown, G., Graves, G., & Ronen, D. (1987). Scheduling ocean transportation of crude
60-day planning horizon. We obtain the optimal solution of 50.4 mu oil. Management Science, 33, 335–346.
within 141 CPU s. However, the model of Christiansen (1999) needs Christiansen, M. (1999). Decomposition of a combined inventory and time con-
1481 CPU s to obtain the same optimal solution. strained ship routing problem. Transportation Science, 33(1), 3–16.
Christiansen, M., Fagerholt, K., & Ronen, D. (2004). Ship routing and scheduling:
We conclude that our model solves faster than the previous Status and perspectives. Transportation Science, 38(1), 1–18.
models in the literature on the above examples and does not violate Dror, M., & Ball, M. (1987). Inventory/routing: Reduction from an annual to a short-
stock limits at any time. period problem. Naval Research Logistics, 34, 891–905.
Fagerholt, K., & Christiansen, M. (2000a). A combined ship scheduling and allocation
problem. Journal of the Operational Research Society, 51, 834–842.
6. Conclusions Fagerholt, K., & Christiansen, M. (2000b). A travelling salesman problem with
allocation, time window and precedence constraints – An application to ship
scheduling. International Transactions in Operational Research, 7, 231–244.
We developed a novel multi-grid MILP formulation based on
Fisher, M, & Rosenwein, M. (1989). An interactive optimization system for bulk-cargo
unit-slots for an important problem faced by most large global ship scheduling. Naval Research Logistics, 36, 27–42.
chemical companies that use or charter a fleet of ships to main- Golden, B., & Wasil, E. (1987). Computerized vehicle routing in the soft drink indus-
tain inventory at many sites globally. The use of multiple slots try. Operations Research, 35, 6–17.
Jetlund, A. S., & Karimi, I. A. (2004). Improving the logistics of multi-compartment
increases the computational efficiency. On four test problems of chemical tankers. Computers and Chemical Engineering, 28, 1267–1283.
varying sizes, our model performs better than that of Li et al. Laporate, G., & Osman, I. H. (1995). Routing problems: A bibliography. Annals of
(2008). Furthermore, in contrast to previous literature, our model Operations Research, 61, 227–262.
Li, J., Karimi, I. A., & Srinivasan, R. (2008). Supply and distribution of multiple mate-
addresses limited jetties, prevents violations of inventory limits, rials via bulk maritime logistics. In 5th foundations of computer-aided process
and most importantly needs no separate estimate of arrivals at operations – FOCAPO Cambridge, Massachusetts, USA, June 29–July 2.
each site. Constant production or consumption rates, determinis- Li, J., Karimi, I. A., & Srinivasan, R. (2010). Recipe determination and scheduling of
gasoline blending operations. AIChE Journal, 56, 441–465.
tic problem parameters, and unlimited raw material availability at Liu, Y., & Karimi, I. A. (2007). Scheduling multistage, multiproduct batch plants
external supplier sites are key assumptions in the model, which with nonidentical parallel units and unlimited intermediate storage. Chemical
require further study to make this model more realistic for real-life Engineering Science, 62, 1549–1566.
Miller, D. M. (1987). An interactive computer-aided ship scheduling problem system.
inventory service problems. While the model applies to any num- European Journal of Operational Research, 32, 363–379.
ber of raw materials, the test problems involved a few materials Papadakis, N., & Perakis, A. (1989). A nonlinear approach to the multiorigin,
and short planning horizons. Further work is warranted to develop multi-destination fleet deployment problem. Naval Research Logistics, 36, 515–
528.
more efficient algorithms that can tackle many more products and
Ronen, D. (1986). Short-term scheduling of vessels for shipping bulk or semi-bulk
longer horizons. commodities originating in a single area. Operations Research, 34, 164–173.
Russel, R. A., & Igo, W. (1979). An assignment routing problem. Networks, 9, 1–
17.
Acknowledgement
Susarla, N., Li, J., & Karimi, I. A. (2010). A novel approach for short-term scheduling
of multipurpose batch plants. AIChE Journal, 56, 1859–1879.
The authors would like to acknowledge financial support for U.S. Department of Transportation, Bureau of Transportation Statistics, Maritime
this work from The Agency for Science, Technology, and Research Administration. (1999). Maritime trade and transportation 1999. Washington, DC:
U.S. Coast Guard.
(A*Star) under grant 052 116 0074.

You might also like