Content uploaded by M. Ali Ülkü
Author content
All content in this area was uploaded by M. Ali Ülkü on Dec 03, 2021
Content may be subject to copyright.
Dare to care: Shipment consolidation reduces not only costs,
but also environmental damage
M. Ali ¨
Ulk ¨
u
n
Capital University, School of Management and Leadership, 1 College and Main, Columbus, OH 43209, USA
article info
Article history:
Received 11 August 2010
Accepted 6 September 2011
Available online 14 September 2011
Keywords:
Environment
Sustainable logistics
Freight transportation
Private carriage
Carbon emissions modeling
Optimization
abstract
Companies interested in greening their supply chains quickly realize that carbon and energy waste
represents an expense, and thus minimizing this expense not only brings in truckloads of money, but
also helps save the environment. This paper demonstrates that shipment consolidation, a powerful
logistics strategy that combines two or more small shipments into an aggregate load to be dispatched
on the same vehicle, can help mitigate carbon and energy waste. Specifically, a Discrete-Time Based
Shipment ConsoLidation (DTB-SCL) policy and a new method to calculate CO
2
emissions associated
with the dispatch of a vehicle are introduced. In addition, an optimization model with the objective of
maximizing combined economic and environmental savings is developed in this study. It is shown that,
among other benefits, the higher use of transportation capacity decreases environmental damage, one
goal of the DTB-SCL policy. The mechanics of the mathematical models are illustrated by numerical
examples, and sensitivity analyses are conducted to provide managerial insights that might help
companies come to better decisions that are environmentally responsible.
&2011 Elsevier B.V. All rights reserved.
1. Introduction
With dramatic changes in technological capacities and a
movement toward globally connected production and distribu-
tion operations in recent decades, alarming environmental issues
have become apparent that require the close and prompt atten-
tion of both practitioners and academicians. Moreover, businesses
should strive to become smarter in their use of resources in their
operations by any means possible, as customers are demanding
more in terms of sustainability.
The environmental damage that accompanies economic develop-
ment calls for serious action from all parties in the market, including
regulatory bodies such as the government, business operators, and
consumers. ‘‘If current predictions of population growth prove
accurate and patterns of human activity on the planet remain
unchanged, science and technology may not be able to prevent either
irreversible degradation of the environment or continued poverty for
much of the world,’’ reports The Royal Society of London and US
National Academy of Sciences (1992).
Accordingly, greening the supply chain has become a major
challenge in business operations. Green Supply Chain Manage-
ment (GSCM) can be defined as ‘‘integrating environmental
thinking into supply chain management, including product
design, material sourcing and selection, manufacturing processes,
delivery of the final product to the consumers as well as end-of-life
management of the product after its useful life’’ (Srivastava,
2007). The smart use of logistics systems has much to offer in
greening the supply chain by means of various efficient distribu-
tion and transportation strategies, especially in the case of the
movement of freight. These strategies that aim to reduce envir-
onmental damage hinge on the efficient use of materials in
production and flow of freight, including purchasing, inbound
logistics, manufacturing, outbound logistics, marketing, after-
sales service and product returns, recycling, remanufacturing,
and centralized distribution (Wu and Dunn, 1995;Beamon,
1999;McKinnon, 2003;Sheu et al., 2005;Aronsson and Brodin,
2006;Sarkis, 2006;Linton et al., 2007;Kohn and Brodin, 2008;
Quariguasi Frota Neo et al., 2008;Allwood et al., 2011).
In logistics systems, transportation is considered to be the
largest source of environmental hazards. Transportation vehicles
not only emit toxic chemicals, but also generate noise pollution. In
the U.S., for example, the predominant greenhouse gas emitted is
carbon dioxide (CO
2
), which accounts for 85% of the climate
change potential for all human-produced emissions. The role of
CO
2
in greenhouse gas emissions accounts for 96% of transporta-
tion-related emissions (EPA, 2006). Thus, CO
2
emissions in the U.S.
are primarily produced by the transportation sector. Emissions
from trucks increased from 42% in 1995 to 49% of total transporta-
tion CO
2
emissions in 2006 and show no signs of decreasing (RITA,
2008). Fig. 1 shows the CO
2
emissions in the U.S. by the end-use
sector from 1990 to 2008. Over the last decade, CO
2
emissions
from the transportation sector have continued to dominate. These
Contents lists available at SciVerse ScienceDirect
journal homepage: www.elsevier.com/locate/ijpe
Int. J. Production Economics
0925-5273/$ - see front matter &2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.ijpe.2011.09.015
n
Tel.: þ1 614 236 6530; fax: þ1 614 236 6540.
E-mail address: aulku@capital.edu
Int. J. Production Economics 139 (2012) 438–446
figures again highlight the crucial importance of mitigating
transportation-related emissions in the context of both public
and private sectors.
Fortunately, some companies have started to realize the value
of greening their operations. For example Wal-Mart, the world’s
biggest retailer, recently undertook a GSCM project and asked its
60,000 suppliers worldwide to reduce their use of packaging by
5%, which amounts to removing 667,000 m
3
of CO
2
from the air
and 213,000 trucks from the road, resulting in a huge savings of
$3.4 billion (Hoffman, 2007).
Motivated by the apparent importance of reducing environmental
damage caused by transportation operations, this research explores
the impact of Shipment ConsoLidation (SCL) on greening and cost-
saving efforts. SCL is a logistics strategy that combines two or more
small shipments on the same vehicle to enable economies of scale.
SCL was shown not only to reduce transportation costs, but, if
employed appropriately, can also prove to be a powerful GSCM
practice.
Fig. 2 presents a graphic view of the problem under discussion.
The customer orders are accumulated at the consolidation facility
of the shipper (e.g., a manufacturer–distributor such as Wal-Mart).
The orders are then transported to the consignee/warehouse and
then locally delivered to retailers in the assigned distribution
zones. These orders are shipped by private carriage, i.e., one’s own
fleet of trucks. In addition, delivery time is guaranteed for each
order. For practicality, the shipper applies a Discrete-Time Based
SCL policy (DTB-SCL) in which consolidated shipments are dis-
patched at specific times of the day or on certain days of the week.
Within this modeling framework, this paper contributes to the
current literature by answering the following research questions:
1. How can the economic and environmental savings that come
from implementing a DTB-SCL policy be modeled analytically?
2. What are the impacts of the key parameters on the perfor-
mance of a DTB-SCL policy?
The organization of this paper is as follows. In the next section,
a background on the uses and mechanics of SCL, along with key
references, are provided. In addition, a brief literature review of
GSCM and the calculation of CO
2
emissions in the context of
transportation are presented. In Section 3, a mathematical model
that quantifies the benefits of DTB-SCL as opposed to immediate
(i.e., without consolidation) shipments is introduced. In Section 4,
a formula to calculate the CO
2
emissions from the dispatch of a
vehicle for a particular lane is proposed. The formula improves on
the current literature by explicitly including the use of the
transportation vehicle by means of both weight and volume
capacities as well as the characteristics of the product shipped.
In Section 5, the utility of the model is further explored through
numerical examples and sensitivity analyses. Finally, discussions
and possibilities for further research are presented.
2. Background on SCL and related literature
SCL is an environmentally responsible logistics strategy that
combines two or more orders (or shipments) so that a larger
quantity can be dispatched on the same vehicle to the same
market region. This can greatly reduce the transportation cost per
item, per order, or per unit of weight. The appropriate use of SCL
helps improve GSCM objectives in terms of less transport effort by
employing fewer long-haul shipments, which results in higher
ton-miles per vehicle per year and lower total vehicle-miles. The
resulting economies of scale in transportation operations make it
possible for shippers and/or carriers to line-haul larger shipments
at lower rates per unit, thereby enabling discount economies to
the customers.
SCL primarily favors the carrier’s pickup, delivery, and dock-
handling costs. For example, a Full-TruckLoad (FTL) shipment
requires only two stops by the carrier: one for pickup at the origin
and one at the destination. By contrast, small shipments require
the Less-Than-TruckLoad (LTL) carrier to make more stops for
pickup and delivery, although with fewer numbers of long-haul
dispatches. Another advantage of SCL is in logistics customer
service; SCL can allow for faster and consistent line-haul or transit
times, which in turn would result in reduced inventories without
changing customer-service standards. Moreover, with faster tran-
sit times, capital is tied up in the consignment for a shorter time,
and fast deliveries might accelerate the cash flow (Masters, 1980;
Cooper, 1984;Hall, 1987;¨
Ulk ¨
u, 2010).
Assuming that delivery-time guarantees are met, SCL can
enable cost savings by allowing a mix of FTL and LTL modes of
shipment. In the FTL mode, the shipper owns or rents the entire
truck space to carry goods directly to customers, whereas, in the
LTL mode, only a fraction of the truck space is paid for. The
shipment starts at a consolidation or distribution center, where
the LTL shipments from the same or different shippers are loaded
on a truck, and ends at the location of a single final consignee.
Naturally, the costing structures for each of these types of
shipments vary. The FTL shipping cost depends on the final
destination and the number of intermediate stops. In contrast,
in LTL operations, the shipping cost per cwt (100 lbs) is mainly
determined based not only on the destination zone, but also on
the weight range of the small shipment. The FTL shipment is cost-
effective if the quantity of freight to be delivered is near the truck
capacity, which can be achieved by SCL (Crainic, 1999;Caputo
et al., 2005;Chopra and Meindl, 2007).
Fig. 3 shows the constraints that are most prevalent in
implementing an SCL program. The appropriate SCL policy for
inbound and/or outbound logistics essentially depends on the
operating environment of the customer order characteristics, such
as product type and due dates as well as the cost and transporta-
tion capabilities of the consolidating party. For example, a shipper
can consolidate its orders going to a specific destination and ship
Fig. 1. CO
2
emissions by end-use sector in the U.S. in 1990–2008.
Source: EIA, 2008.
Fig. 2. Schematization of the shipping and distribution problem.
M.A. ¨
Ulk¨
u / Int. J. Production Economics 139 (2012) 438–446 439
them to the consignee using its own fleet. Alternatively, a carrier
can consolidate orders from different shippers at a make-bulk
terminal, line-haul a particular lane, and then break-bulk the
consolidated load at the destination terminal for local deliveries
to individual customers. The latter operation uses what is termed
common carriage, i.e., a for-hire trucking company.
The fundamental decision in developing an SCL policy is when
to ship. In the model formulation for quantifying the savings by
employing SCL because of its ease of implementation, this research
particularly develops a DTB-SCL policy in which each order
(whether consolidated or not) is dispatched on a predetermined
shipping date. Rather than being continuous, the length of the
DTB-SCL cycle is in multiples of time periods such as half days.
(The details of this model are given in Section 3. For analytical cost
comparison and for simulation-based environmental assessment
of the other commonly used SCL policies, see ¨
Ulk ¨
u (2009) and
Merrick and Bookbinder (2010) for examples.) The following is a
brief review of recent studies addressing the calculation of GSCM
and CO
2
emissions.
The body of literature on quantifying environmental damage
emitted by supply chain operations is not exhaustive. In one of
the earlier studies, Van Hoek (1999) suggested that, on the supply
chain level, performance measures, such as emission rates and
energy efficiency per material, would be appropriate. Murphy and
Poist (2003) found that U.S. and non-U.S. firms have similar
perspectives on most of the issues regarding the environmental
management of logistics. Yang et al. (2005) examined the envir-
onmental issues for a postponement strategy that proposes
transportation as a mediating factor and present possible ways
to mitigate the effects of increased traffic congestion and pollu-
tion. Ferretti et al. (2007) investigated the integration of causes of
transportation pollution by drawing on an industrial case study
for an aluminum supply chain and offering an analytical model to
evaluate the economic and environmental effects of such integra-
tion. In Sbihi and Eglese (2007), various mathematical models
are provided to optimize system-wide performance when the
objectives comprise not only economic, but also environmental
and social considerations. Sundarakani et al. (2010) proposed a
detailed diffusion model to estimate carbon footprints, implying
that carbon emissions across stages in a supply chain can
constitute a significant threat that warrants careful attention in
the design phase of supply chains. In a recent study by Benjaafar
et al. (2010), carbon footprint parameters were gauged with
the simple decision variables used primarily in procurement,
production and inventory management, whereas Ubeda et al.
(2010) showed in a case study how companies can turn their
practices green while satisfying efficiency goals. Sarkis et al.
(2011) surveyed the GSCM literature from an organizational and
theoretical perspective and suggested research opportunities and
directions.
As noted in Section 1, trucking accounts for the majority of
freight transportation, and it is therefore included in the model
presented in this paper. Although a number of new studies have
been conducted on incorporating CO
2
emissions into decision
models at the supply chain level, as pointed out by McKinnon
(2000), further research is required for more detailed CO
2
emis-
sions calculations, such as the consideration of the impact of the
vehicle load factor on environmental pollution. In that study,
McKinnon surveyed opportunities to improve the use of road
vehicles and suggested that the impact of freight transportation
on the environment can be reduced ‘‘by increasing the share of
freight moved in less environmentally damaging ways, by
increasing vehicle load factors, by improving freight transport
operations’ energy efficiency, and by reducing emissions per unit
energy consumed in freight transport.’’ He reports that in coun-
tries where the road is the dominant means of transportation,
such as the U.S. and Canada, increasing truck loading could yield
greater environmental benefits than a modal shift, for example,
from trucking to rail. He also showed that increasing the max-
imum truck weight can further yield economic and environmen-
tal benefits (McKinnon, 2005). Samuelsson and Tilanus (1997)
investigated the general physical efficiencies of goods transporta-
tion; however, they do not specifically include load factors
in their models. In a recent paper, Figliozzi (2011) quantified
CO
2
emissions emanating from urban freight distribution in the
presence of traffic congestion during peak hours.
3. Modeling and analysis of the DTB-SCL policy
For the modeling purposes of the DTB-SCL policy, consider a one-
tiered supply chain that consists of one shipper and one consignee.
The transportation of freight such as a particular consumer good is
performed by the shipper’s private fleet of vehicles. The objective is
to find the length of the DTB-SCL cycle Tthat enables the most
savings in both cost and carbon emissions. Assume that the orders
received are unit-sized but with no restriction on their size relative
to the vehicle capacity. Orders larger than the capacity of a single
vehicle are allowed to be split among multiple vehicles, as long as
those partial deliveries are made within the guaranteed delivery
time. This is a plausible proposition and is indeed a practice
implemented by distributors such as Amazon.com. In this section,
without any loss of generality, the vehicle capacity is considered to
be the weight capacity. The road vehicles (e.g., trucks) employed in
the shipper’s fleet are all of the same type, and it is assumed that the
shipper has an ample number of them so that there is no need for
outsourcing extra transportation capacity.
The DTB-SCL policy builds on the practice of allowing the
shipment releases only at certain discrete times of the day, such
as noon, evening, or midnight. The length of the time period (e.g.,
6 h or a full day) is a managerial decision and depends on the
planning horizon of the logistics company. To aid the reader, the
definitions of the decision variables and parameters are given in
Table 1. Other less frequently used notations will be described in
each relevant section of this paper.
Building on the DTB-SCL policy, this paper develops two optimi-
zation models. The economic-savings model (Model DTB-SCL-I)
focuses on maximizing cost savings, whereas the Model DTB-SCL-II,
which will be introduced in Section 4, integrates the minimization of
CO
2
emissions with Model DTB-SCL-I. The explanation and formula-
tion of Model DTB-SCL-I are given by the following:
Proposition 1. Consider a sequence of shipments of a single type of
product over an interval of T periods by a private fleet that can
supply multiple trucks of the same type as required. (The product and
vehicle indices k and p in the relevant parameters are suppressed.) An
Fig. 3. Policy variables for SCL programs.
Source: Adapted from Higginson and Bookbinder, 1994.
M.A. ¨
Ulk¨
u / Int. J. Production Economics 139 (2012) 438–446440
‘‘immediate shipment’’ policy dispatches a vehicle each period (often
containing a partial load) as long as there is a demand in the
corresponding period, whereas the DTB-SCL policy waits and releases
a full vehicle one or more times during the T-period consolidation
cycle. If necessary, a partially loaded truck is dispatched at the end of
this interval. Determining the optimal length of the DTB-SCL cycle T
n
can then be cast as the following nonlinear integer optimization
problem:
Economic Savings Optimization Model (DTB-SCL-I)
T
n
¼arg maxfTC
I
ðTÞTC
C
ðTÞg such that ð1Þ
TrT
max
¼GLZð2Þ
TZT
min
¼V
VA
&’ ð3Þ
N
I
¼A
V
ð4Þ
TC
I
ðTÞ¼TFN
I
þC
2
V
2
ðN
I
1Þ
AþðAVðN
I
1ÞÞ 1VðN
I
1Þ
A
ð5Þ
N
C
¼AT
V
ð6Þ
TC
C
ðTÞ¼FN
C
þC
2
V
2
ðN
C
1Þ
AþðATVðN
C
1ÞÞ TVðN
C
1Þ
A
ð7Þ
TA½T
min
,...,T
max
Z
þ
ð8Þ
Proof. Eq. (1) asserts that the optimal cycle length T
*
is the one that
yields the maximum cost savings per cycle when the DTB-SCL policy
is used as opposed to the immediate shipment policy. Note that
there needs to be an upper bound T
max
on the length of the DTB-SCL
cycle Tso that a certain customer service level is achieved. That
upper bound, which is developed as a linear function of the delivery
time guarantee G, the line-haul time L, and the local delivery time Z,
isgiveninEq.(2).Theobjectivevalue,i.e.,thecostsavingspercycle,
in Eq. (1) will be positive only if Tis at least as great as the length of
time required to receive enough orders so that a residual vehicle
capacity is efficiently utilized. This lower bound T
min
amounts to the
least number of periods that is greater than the ratio of the vehicle
capacity Vto the absolute difference between the order arrival rate A
and the vehicle capacity; cf. Eq. (3). &
The fixed cost Fof dispatching a vehicle is independent of the
quantity of freight shipped. However, if the accumulated load is more
than a single vehicle’s capacity V, then the quantity remaining is to be
shipped at the end of the period on another vehicle, which is only
partially loaded. The total cost per cycle would be the sum of the
consolidation and dispatch costs over a cycle. Let the total number of
vehicles employed in each period be N
I
, where the subscript Istands
for ‘‘immediate shipment’’ policy. Then N
I
can be calculated using
Eq. (4). Without loss of generality, suppose there are (N
I
1) fully
loaded vehicles and one partially loaded vehicle. The time required to
fill a vehicle is V/Aperiods. Naturally, in a period, if enough orders are
accumulated before the end of that period, a vehicle is dispatched as
soon as it is fully loaded. For such an order, the average wait time
then is V/2Aperiods. Then the consolidation cost for a single FTL is
CV(V/2A), and the total consolidation cost for all of the FTL shipments
in a period is simply CV
2
((N
I
1)/2A). The total amount of the last
(possibly partial) load that is to be dispatched at the end of that
period is now AV(N
I
1) orders. Hence, the average waiting time for
an order that is shipped in a partially loaded truck is (1V(D
I
1)/A)/
2 periods. Therefore, the consolidation cost for a partial load in a given
period is (C/2)(AV(N
I
1))(1V(D
I
1)/A) dollars. Including the
fixed cost of dispatching D
I
vehicles in each of Tperiods, the total
cost of immediate shipments TC
I
(T) is obtained by Eq. (5).
Now, let TC
C
(T) be the cost of implementing the DTB-SCL policy
over Tperiods. (The subscript C will refer to ‘‘consolidation.’’) In
such a policy, a vehicle will be held until enough orders are received
to fill it. Obviously, if the order arrival rate is less than the vehicle
capacity, a vehicle will be held for more than one period. Then the
number of vehicles required to dispatch the load consolidated over
Tperiods, N
C
, is calculated by Eq. (6). Note the distinction that the
total number of dispatches in immediate shipments during T
periods is TN
I
, whereas it is only N
C
when the DTB-SCL policy is
employed. In this case, there will definitely be (N
C
1) fully loaded
vehicles and possibly one partially loaded vehicle. Similar to that of
TC
I
(T), TC
C
(T) can be derived as in Eq. (7). The difference between
TC
I
(T)andTC
C
(T) gives the desired objective function. Finally, Eq. (8)
implies that the optimization search needs to be over the positive
integers inclusively between the lower and upper bounds of the
length of the DTB-SCL cycle T. This completes the proof.
Proposition 1 provides the conditions under which the
DTB-SCL policy provides lower costs, which is T
max
ZTZmaxfT
min
,
dargfTC
I
ðTÞTC
C
ðTÞZ0eg. Specifically, the optimization of economic
savings is given by Model DTB-SCL-I. Carrying the same quantity of
freight with a fewer number of dispatches certainly reduces carbon
emissions and noise pollution ceteris paribus. Accordingly, how can
the amounts of CO
2
emitted by the dispatch of a freight transporta-
tion vehicle be calculated? How can the environmental savings then
be incorporated into Model DTB-SCL-I? Section 4 will examine these
questions.
4. Calculating the amount of CO
2
emissions from a dispatch
Increasing truck loading, in turn, implies increasing the vehicle
load factor or vehicle utilization, which is expressed as the ratio of the
actual weight of goods carried to the maximum weight that could
Table 1
Model Nomenclature.
Symbol Description Measure
TLength of the DTB-SCL cycle time periods
FFixed cost of dispatching a vehicle for a
particular lane
$/dispatch
CConsolidation cost per order per period $/order/period
A
p
Order arrival rate for product type-p orders/period
V
k
Volume capacity of vehicle type-k cubic feet (ft
3
)
W
k
Weight capacity of vehicle type-k cwt (100 lbs)
M
k
Average mileage of vehicle type-k miles/gal
F
CO
2
k
Fuel emissions factor lbs/gal
S
CO
2
k,p
Shipping emissions factor lbs/cwt-mile
P
CO
2
k,p
Packing emissions factor lbs/order
g
k,p
Packing efficiency factor (0r
g
k,p
r1) dimensionless
v
p
Volume of a unit of product type-p ft
3
w
p
Weight of a unit of product type-p cwt
y
v
Target volume efficiency level (0r
y
v
r1) dimensionless
y
w
Target weight efficiency level (0r
y
w
r1) dimensionless
r
Utilization of vehicle capacity (0r
r
r1) dimensionless
ZAverage local delivery time for an order time periods
LLine-haul (transit) time time periods
DLine-haul distance miles
c
f
Congestion factor for a particular line-haul
lane (0rc
f
o1)
dimensionless
GDelivery time guarantee, a customer service
measure
time periods
T
max
Maximum admissible length of the DTB-SCL
cycle
time periods
M.A. ¨
Ulk¨
u / Int. J. Production Economics 139 (2012) 438–446 441
have been carried on a laden trip. Vehicle utilization can be improved
by SCL.
In the literature, the calculation of CO
2
emissions from trans-
portation essentially is based on the weight of the load, the vehicle
and type of fuel used, and the distance elapsed to carry that load.
For example, the Transport Research Laboratory (TRL, 1999)has
developed the following function through statistical analysis that
links the emissions to the travel speed and distance: dð
a
0
þ
a
1
sþ
a
2
s
3
þ
a
3
s
2
Þ,wheredand sare the distance and average speed for
a particular origin-destination, and where the remainder of the terms
are several constants specified according to the type of vehicle.
For heavy-duty trucks, for example, those constants are provided as
(
a
0
,
a
1
,
a
2
,
a
3
)¼(1576.0, 17.6, 0.00117, 36067.0). Although this and
other similar CO
2
calculations are easy to implement, they fail to
account for the fact that low-density (low weight-to-volume ratio)
products may cube out long before the maximum permitted weight
range is achieved. Hence, a more realistic approach should include
not only the specifics of the products shipped, but also how
efficiently they are packed in the vehicle.
To that end, a novel calculation of CO
2
emissions is proposed that
integrates both weight and volume efficiencies. To lay the foundation
for this computation, let us define the ‘‘shipping’’ emission factor S
CO
2
k,p
as the average amount of CO
2
per mile emitted when vehicle type k
carries a hundredweight (i.e., cwt) of product type p. Let us also
define the ‘‘packing’’ emission factor P
CO
2
k,p
as the average amount
of CO
2
emissions per packaging of a standard unit of product type
p, including its loading and unloading into vehicle type k.The
shipping emission factors for different modes of transportation are
already available (e.g., www.carbonfund.org). For example, the
amounts of CO
2
emissions per cwt-mile are 0.0805, 0.0169, 0.0105,
and 0.0040 lbs for air, road, rail, and sea transportation, respectively.
However, the packing emissions factor as defined in this paper is
specific to both the product type and the vehicle type. (Additional
research is required on carbon emissions attributable not only to
transportation, but also to product-specific consumer packaging.
Hekkert et al. (2000) showed that the CO
2
emissions related to the
production and use of transport packaging can be reduced up
to 40%.).
Consider the dispatch of a vehicle type kwith weight capacity
W
k
, volume capacity V
k
, and an ‘‘actual’’ average speed of S
k
for a
particular line-haul distance of Dmiles on which the maximum
speed allowed is S
max
. Considering that S
k
oS
max
, the traffic
congestion factor c
f
can be derived as a function of the ratio of
these two speed values. Assume that the freight transported is of
product type p occupying a space of v
p
cubic feet and weighing w
p
cwt. Let
y
v
and
y
w
be the volume and weight efficiency levels set
as targets by the logistics manager, respectively. For example,
y
w
¼0.85 would allow the dispatcher to load up to 85% of the
weight capacity of the vehicle. Let
g
kp
be the ‘‘packing’’ efficiency,
which measures how well type-p products can be packed into
vehicle type k. This is an important factor to consider because the
ways that products can be stacked vary, which, in turn, affects the
utilization of the vehicle capacity. Finally, let the fuel emissions
factor F
CO
2
k
be the amount (lbs) of CO
2
emissions per gallon of fuel
used to run type-k vehicle, and let M
k
be the average mileage
(miles/gal). Then the amount of CO
2
emitted from the Dmiles
line-haul of a type-k vehicle loaded with type-p product E
CO
2
k,p
ðDÞ
can be approximated by the following:
E
CO
2
k,p
ðDÞ¼F
CO
2
k
D
M
k
þP
CO
2
k,p
U
k,p
þS
CO
2
k,p
U
k,p
Dw
p
ð1c
f
Þwhere
U
k,p
¼min V
k
g
k,p
y
v
v
p
,W
k
y
v
w
p
,and c
f
¼1S
k
S
max
:ð9Þ
The variable U
kp
determines the maximum number of type-p
products that can fit in a single type-k vehicle. The formulae in
Equation set (9) are easy to execute and offer logistics managers
important opportunities for improving efficiency in their environ-
mentally responsible shipment-release decisions. Therefore, consider
the following:
Illustrative Example: Assume that a particular type of transporta-
tion vehicle is to be packed with a specific product. (Accordingly,
the vehicle and product indices k and p will be dropped from the
relevant expressions.) A specific bookcase, manufactured and dis-
tributed by Company XYZ, is high in demand. Naturally, the
company wants to ship as many bookcases as possible from its
consolidation site to its distribution center, which is D¼1000 miles
away. That bookcase weighs w¼1.5 cwt and occupies a volume of
v¼20 ft
3
. Company XYZ owns only one transportation vehicle, a
14-foot truck with weight capacity W¼65 cwt and volume capacity
V¼1400 ft
3
and that runs an average mileage of M¼10 miles/h.
When hauling back empty, the truck emits F
CO
2
¼19:4lbs of CO
2
per gallon of gasoline (www.epa.gov). The shipping emissions factor
S
CO
2
for the freight on this type of truck is estimated to be
0.0169 lbs of CO
2
per cwt mile (www.carbonfund.org). The manu-
facture of one lb of a corrugated box emits approximately 1.6960
lbs of CO
2
(www.fefco.org). Assuming that the amount of the
corrugated box used to wrap a bookcase weighs 1.3 lbs, the packing
emissions factor P
CO
2
is 2.2048 lbs of CO
2
per bookcase, i.e., per
order. In addition, assume that although the bookcases can be
stacked in the truck on any side of the box, the geometric shape of
the truck allows for only 90% packing efficiency (
g
¼0.9). Moreover,
the logistics manager desires 85% target weight and 100% target
volume efficiencies;
y
w
¼0.85 and
y
v
¼1. The actual average speed
of this truck on the aforementioned long-haul lane is estimated
to be 57 miles/hr. The maximum allowable speed is 65 miles/h.
Then the traffic congestion factor is c
f
¼1(57/65)¼0.1231. This
truck can be loaded with at most U¼minfbðð1400Þð0:9Þð1ÞÞ=ð20Þc,
bðð65Þð0:85ÞÞ=ð1:5þ0:013Þcg ¼ minf63,36g¼36 such bookcases.
After filling in the relevant parameters, the total amount of CO
2
emissionsfromthisdispatchis
E
CO
2
ð1000Þ¼ð19:6Þ1000
9þð2:2048Þð36Þþð0:0169Þð36Þð1000Þð1:513Þ
ð10:1231Þ
¼2928:7lbs:
Eq. (9) explicitly incorporates the vehicle and product type
characteristics, distance elapsed, mileage, fuel type, traffic congestion
effect, and effective volume and weight utilizations in finding the
CO
2
emissions per dispatch. For further insights through structural
properties of the formulation in Eq. (9), two more definitions are
needed. Let n
k,p
be the ‘‘actual’’ average number of type-p orders
carried in a type-k vehicle over the long run. (Note that n
k,p
rU
k,p
and that n
k,p
is related to the actual order arrival rate.) Now, define
the ‘‘utilization of vehicle capacity’’ by
r
n
k,p
=U
k,p
.
Proposition 2. The amount of CO
2
emissions per order for a vehicle
of type k that line-hauls on average n
k,p
type-p orders over a distance
of D miles, E
CO
2
k,p
ðD,n
k,p
Þ
(a) increases in the line-haul distance D, the traffic congestion factor
c
f
, the fuel emissions factor F
CO
2
k
, the packing emissions factor
P
CO
2
k,p
, and the shipping emissions factor S
CO
2
k,p
and
(b) decreases in the utilization of vehicle capacity
r
, the target
volume efficiency
y
v
, the target weight efficiency
y
w
, the mileage
of the vehicle M
k
, and the packing efficiency factor
g
kp
.
Proof. These claims can be justified by simply checking the
definitions of the variables in Eq. (9) and by noting that
E
CO
2
k,p
ðD,n
k,p
Þ¼ E
CO
2
k,p
ðDÞ
n
k,p
¼F
CO
2
k
D
r
M
k
U
k,p
þS
CO
2
k,p
Dw
p
ð1c
f
ÞþP
CO
2
k,p
:&
M.A. ¨
Ulk¨
u / Int. J. Production Economics 139 (2012) 438–446442
Proposition 2 aligns with intuition and provides insights into how
to reduce emissions in freight transportation as well as analytically
shows that increasing the utilization of the vehicle capacity reduces
emissions. Proposition 2 also states that as the packing efficiency
factor
g
kp
is improved, the effective volume utilization increases,
which in turn decreases the amount of CO
2
emissions. Hence,
management should consider ways to stack the freight in a vehicle
such that the effective capacity is increased. This might be achieved,
for example, by incentivizing the manufacturer to use green and
smart consumer packaging, using recyclable and modular packaging,
by improving the efficiency of packing, and by optimization of vehicle
routing. Poor packaging not only creates emissions at the manufac-
turing stage, but also creates difficulties in loading, unloading, and
sorting operations at the warehouses or retailers. Light, dense, and
modularpackaging(apracticeledbyIKEA)savescostsandcarbon
emissions. Moreover, the choice of a greener transportation vehicle
with less fuel emissions and with higher mileage is important.
Although they are implicit, several other results can be
obtained from Proposition 2. For example, for a given quantity
of freight transported on a specific lane, the carbon emissions can
be reduced if the vehicle capacity is increased; cf. McKinnon
(2005). If the speed limit of that specific line-haul is increased or
congestion is reduced, as mentioned, there will be less environ-
mental damage. However, these issues may require reinvestiga-
tion of government regulations, improvements in transportation
infrastructure, and most importantly, a mechanism to coordinate
supply chain parties in greening their logistics operations.
TheessenceofProposition 2 is that the utilization of vehicle
capacity needs to be increased. The DTB-SCL policy serves exactly that
purpose because it aggregates the orders over time and then ships
them with fewer and fuller vehicles. Hence, Eq. (9) provides a means
to calculate the environmental savings that come with the DTB-SCL
policy. Recalling Proposition 1, the number of line-haul dispatches
saved by applying Model DTB-SCL-I is simply TD
I
D
C
.Then,the
combined economic–environmental savings under the DTB-SCL policy,
which will be called Model DTB-SCL-II, can be constructed as a multi-
criteria nonlinear integer optimization model with the following
modifications:
Economic–Environmental Savings Optimization Model
(DTB-SCL-II):
max½TC
I
ðTÞTC
C
ðTÞþ
b
E
CO
2
k,p
ðDÞ½TN
I
N
C
subject to ð10Þ
V¼Uv
p
ð11Þ
L¼D
S
k
ð12Þ
and constraints (2)–(8).
In Eq. (10),
b
is the cost of emitting one lb of CO
2
emissions, a
cost figure that is calibrated by management. Note that if
b
is set
to zero, Model DTB-SCL-II reduces to Model DTB-SCL-I, in which
environmental savings are achieved, although not necessarily
optimally.
5. Numerical examples and sensitivity analyses
The computational mechanics of the DTB-SCL policy and
further insights into the approach can be obtained by conducting
numerical examples and sensitivity analyses. For these purposes,
the ‘‘base-case’’ parameter values chosen are given in Table 2.
What is the optimal length of the DTB-SCL cycle that yields the
maximum economic and/or environmental savings? Using the
base case data, the lower and upper bounds on the decision
variable are computed as T
min
¼10=ð107Þ
¼4 and T
max
¼20
102¼8 periods. The cost calculations are straightforward using
the formulas derived in Sections 3 and 4. For instance, the cost of the
immediate shipment policy for 5 periods is TC
I
(5)¼$587.5, in which
N
I
¼d7=10e¼1 vehicle is dispatched in each period. In contrast, the
DTB-SCL policy requires dispatching of N
C
¼d35=10e¼4vehicles
during each 5-period DTB-SCL cycle, resulting in a total cost of
TC
C
(5)¼$516.1. Then the cost savings per cycle can simply be
computed as TC
I
(T¼5)TC
C
(T¼5)¼587.5 516.1 ¼$71.4 with one
less line-haul dispatch.
Model DTB-SCL-I is aimed at maximizing the savings per cycle.
However, the management might prefer a particular objective to
others, and thus an examination of their relative impacts on the
optimal results and other commonly used objectives in SCL
decisions are also included. These objective functions may consist
of maximizing the savings per order ½TC
I
ðTÞTC
C
ðTÞ=AT, the
savings per period ½TC
I
ðTÞTC
C
ðTÞ=T, the percentage of savings
per cycle [TC
I
(T)TC
C
(T)] 100/TC
I
(T), and the percentage of
savings per period ½TC
I
ðTÞTC
C
ðTÞ 100=½TC
I
ðTÞTand reducing
line-haul dispatches [TD
I
(T)D
C
(T)]. Table 3 displays the corre-
sponding results for each of these objective functions. These
optimal objective values are found through enumeration over
the feasible set of the lengths of the DTB-SCL cycle, i.e., TA[4,...,8].
Because the DTB-SCL savings model is highly nonlinear and
discontinuous and because, in practice, the number of feasible T
values is not large, optimization by exhaustive enumeration is
plausible. (For a continuous time SCL policy combined with pricing
decisions, ¨
Ulk ¨
u and Bookbinder (forthcoming) find the exact
analytical solution for the optimal length of the consolidation
cycle.) For the base case example, the corresponding optimal
objective values are in bold type. Note that with respect to all but
the objective of percentage savings per period, the optimal length
of the DTB-SCL cycle is 7 periods. This small numeric example
Table 2
Base-case parameter values.
Parameters ACF G L V Z
Value 7 5 100 20 10 10 2
Table 3
Optimal results for various objective functions.
TSavings per
cycle
Savings per
order
Savings per
period
% Savings per
cycle
% Savings per
period
Reduction in
line-hauls
4 75.7 2.7 18.9 16.1 4.0 1
5 71.4 2.0 14.3 12.2 2.4 1
6 60.7 1.4 10.1 8.6 1.4 1
7150.7 3.1 21.5 18.3 2.6 2
8 148.6 2.7 18.6 15.8 2.0 2
M.A. ¨
Ulk¨
u / Int. J. Production Economics 139 (2012) 438–446 443
shows the impact of using various objective functions. With
respect to Model DTB-SCL-II and the fact that
b
40, the optimal
length of the DTB-SCL cycle that yields the maximum cost and
environmental savings value of 150:7þ2
b
E
CO
2
k,p
is 7 periods.
In the remainder of this section, the performance of the
DTB-SCL policy will be investigated by sensitizing a key para-
meter ceteris paribus. First, the impact of the cost ratio F/Con the
percentage of cost savings is explored in Fig. 4.F/Cis the ratio of
the cost of dispatching a vehicle to the cost of consolidating an
order for 1 period. Keeping F¼100 and for varying values of C,F/C
values in the range of 10–100 are chosen. From Fig. 4, regardless of
the value of the cost ratio, the percentage of savings per cycle is
nonlinear in the length of the consolidation cycle, but with an
increasing trend and an optimal length of 7 periods of the DTB-SCL
cycle. Another immediate observation is that the higher the cost
ratio, the higher are the percentages of savings. Although not
trivial, this graph provides another insight, which is that the
logistics managers should strive to increase the allowable time-
window (T
max
) by reducing the line-haul time (e.g., by choosing
the shortest-time route) and by minimizing the local pickup and
delivery times (e.g., by solving a vehicle-routing optimization
problem), if possible. Finally, note that the DTB-SCL policy yields
at least 10% and up to nearly 26% in cost savings.
To explore how demand uncertainty affects the DTB-SCL
policy, let us define the freight demand density A/Vas the ratio
of the order arrival rate to the vehicle capacity. Fig. 5 illustrates
the relationship between the savings per order and the length of
the consolidation cycle for varying freight demand densities
ranging from 0.1 to 1.6 in increments of 0.3, in which the vehicle
capacity is set to a fixed value of 10 orders. First, note that there is
no savings if the freight demand density is 1, i.e., when the orders
are equal to the size of the vehicle. This implies that if the order
sizes are multiples of any vehicle capacity, there is no benefit in
consolidating shipments and that the orders should be shipped at
the end of the period in which they arrived. Second, note that the
savings per order are not necessarily linear in the length of the
DTB-SCL cycle. The highest savings per order is achieved when the
freight demand density is very low. When A/V¼0.4, for example,
the optimal consolidation cycle length T
*
is 5 periods. Overall, this
sensitivity analysis shows that implementing the DTB-SCL policy
can enable up to $70 savings per order, which is 70% of the cost.
The preceding sensitivity graphs (Figs. 4 and 5) focus on the
economic impact of SCL. In the remainder of the paper, Figs. 6–8
will provide information about the environmental impact of
applying the DTB-SCL as opposed to the immediate shipment
policy.
As shown with Proposition 2, the amount of CO
2
emissions per
order transported decreases with the vehicle capacity utilization.
Thus, it is of interest to examine how this utilization is affected by
the length of the consolidation cycle and by variations in order
arrival rates. Let us define the ‘‘average’’ utilization of vehicle
capacity as the total load transported divided by the total number
of vehicles dispatched in a cycle. Fig. 6 demonstrates the relation-
ship between the average utilization of vehicle capacity as the
freight demand density varies over the range of 0.1–1.6 in
increments of 0.3, particularly when the DTB-SCL cycle length is
4 periods. Except for the case of FTL operations (i.e., when A/V¼1),
the average utilization of vehicle capacity for the DTB-SCL policy
Fig. 4. Impact of Ton the percentage of savings per cycle for F/C¼10–100.
Fig. 5. Impact of Ton the savings per order for A/V¼0.1–1.6.
Fig. 6. The average utilization of vehicle capacity versus freight demand density
A/Vfor T¼4.
Fig. 7. The number of line-haul dispatches reduced when using DTB-SCL versus
freight demand density A/Vfor T¼4–8.
M.A. ¨
Ulk¨
u / Int. J. Production Economics 139 (2012) 438–446444
is higher than that for the immediate shipment policy. Therefore,
DTB-SCL is proven to aid in protecting the environment by
reducing carbon emissions. Another benefit of SCL is that, for a
given total amount of freight, the reduced number of dispatches
implies less traffic congestion, which, in turn, lowers the CO
2
emissions per dispatch (in Section 4). It should also be noted that
a separate analysis justifies that the findings above are valid not
only for T¼4, but also for the other feasible values 5–8.
The relationship between the number of line-hauls (i.e., the
number of vehicles dispatched) reduced per cycle and the freight
demand density is shown in Fig. 7. The figures are shown for all of
the admissible lengths of the DTB-SCL cycle, i.e., for 4,y,8 periods
and for the freight demand density in the interval of 0.1–1.6.
Unless the freight demand density is 1, the DTB-SCL policy
reduces the number of line-hauls required. This reduction is
achieved mostly when the freight demand density is low. (The
approximate amount of CO
2
emissions reduced by dispatching
one less vehicle E
CO
2
k,p
can be calculated using the method proposed
and as implemented in the numerical example in Section 4.)
Recall that once a test bed is conducted to determine the
parameter values required for computing the carbon emissions
E
CO
2
k,p
ðDÞfor a particular product and a vehicle type and for a
specific line-haul of distance D, the total carbon savings in SCL can
be more accurately computed as E
CO
2
k,p
ðDÞ½TD
I
D
C
.
Depending on the operating environment of the supply chain
as well as the demand characteristics of orders for the products
(recall Fig. 3), consolidating loads of shipments might not be at all
beneficial to the shipper or the carrier. This is especially true if the
orders are very time sensitive (e.g., express delivery) or the
product is perishable (e.g., blood, some pharmaceuticals, or fresh
produce). The models developed in this paper account for this
important consideration by incorporating the delivery time guar-
antee Gas a model parameter. Gaffects the upper bound on
the length of the DTB-SCL cycle T
max
; cf. Eq. (2). Naturally, the
delivery time guarantee influences customer demand and thereby
the order arrival rate. Assuming that the average local delivery
time for an order Zand the line-haul time Lare constant, an
increase in Gimplies an allowable increase in T
max
.
For the values of freight demand densities (i.e., 0.1, 0.4, 0.7, 1, 1.3,
and 1.6) and for Gtaking on values over the range [14,30] with
increments of 2 periods, Fig. 8 shows the number of line-haul
dispatches reduced by the use of the DTB-SCL policy. As Gincreases,
except for FTL-type orders (i.e., when A/V¼1), the reduction in the
number of line-haul dispatches increases. This translates into a
substantial decrease in environmental damage. However, one should
be cautious about customer service considerations; in practice, there
is a maximum delivery time guarantee above which customers
might not accept. Analogous to the behavior exhibited in Fig. 8,it
is correct to state that, if Gis taken as fixed, then any reduction in
either Zor Lor both would yield higher cost savings with lower CO
2
emissions.
6. Concluding remarks
In an effort toward promoting greener supply chains, this
paper introduces the DTB-SCL models to quantify the benefits of
consolidating shipments in terms of both cost and CO
2
emissions
savings in long-haul transportation to improve the sustainability
of logistics systems. In addition, a formula is presented to
calculate the carbon emissions per dispatch of a vehicle, many
of which can be eliminated by the appropriate use of the DTB-SCL
policy. The problem introduced in this paper has many implica-
tions and applications on sustainable logistics supply chains.
There is a great potential to explore some challenging problems
with a direct impact on societal issues that call for prompt
attention, including those that will appear in the carbon trading
market in the near future. Drawing on the limitations of the
models constructed in this paper, below are some future research
topics.
It will be particularly interesting to examine the impact of SCL
on multi-echelon or global logistics operations. Developing a
large-scale mathematical model that explicitly incorporates logis-
tics service levels across more closely intertwined supply chain
configurations is an especially challenging problem because
poorly designed and implemented SCL programs can hurt delivery
time guarantees.
The standardization of product packaging enables streamlined
loading and unloading of the vehicle and improves transportation
capacity utilization. In practice, the most use of SCL involves
aggregating many small orders with different shipment charac-
teristics that arrive randomly. Hence, further research on how to
efficiently consolidate and to pack orders into a transportation
vehicle are required.
The optimization models in this paper assume the use of a fleet
of similar transportation vehicles by a private carrier. Although
this assumption enables analytical tractability, many logistics
companies own or lease a variety of vehicle types, which suggests
a new research topic in which decisions about fleet size need to
be integrated. Moreover, from a global perspective, the possibility
of using different modes of transportation within and between
continents offers another interesting topic for future study when
designing SCL policies with the objective of optimizing combined
economic–environmental efficiency.
In conclusion, this research invites compassionate academi-
cians to invest additional research efforts into the prevailing
needs of the world that necessitate better creation and allocation
of sustainable resources and asks practitioners to ‘‘dare to care,’’
not only about the financial bottom line, but also about the
societal and environmental consequences of their business opera-
tions. To that end, using environmentally responsible decision
mechanisms such as the DTB-SCL policy represents a good start.
Acknowledgments
I am grateful to the guest editor, Dr. J. Sarkis, and the anonymous
referees whose constructive suggestions substantially improved
the clarity of this manuscript. I also thank Dr. J. H. Bookbinder and
Dr. M. A. Turnquist for their insightful comments on an earlier
version of this manuscript. This research has been partially funded
by National Science Foundation, Grant DUE0618252.
Fig. 8. The number of line-haul dispatches reduced when using DTB-SCL versus
delivery time guarantee Gfor freight demand densities A/V¼0.1–1.6.
M.A. ¨
Ulk¨
u / Int. J. Production Economics 139 (2012) 438–446 445
References
Allwood, J.M., Ashby, M.F., Gutowski, T.G., Worrell, E., 2011. Material efficiency: a
white paper. Resources Conservation and Recycling 55 (3), 362–381.
Aronsson, H., Brodin, M.H., 2006. The environmental impact of changing logistics
structures. The International Journal of Logistics Management 17 (3), 394–415.
Beamon, M.B., 1999. Designing the green supply chain. Logistics Information
Management 12 (4), 332–342.
Benjaafar, S., Li, Y., Daskin, M., 2010. CarbonFootprint and the Management of Supply
Chains: Insights from Simple Models. Working Paper.
Caputo, A.C., Fratocchi, L., Pelagagge, P.M., 2005. A framework for analyzing long-
range direct shipping logistics. Industrial Management & Data Systems 105 (7),
876–899.
Chopra, S., Meindl, P., 2007. Supply Chain Management: Strategy, Planning, &
Operation. Prentice Hall, NJ.
Cooper, M.C., 1984. Cost and delivery time implications of freight consolidation
and warehousing strategies. International Journal of Physical Distribution &
Materials Management 14 (6), 47–67.
Crainic, T.G., 1999. Long-haul freight transportation. In: Hall, R.W. (Ed.), Handbook
of Transportation Science, Kluwer, Dordrecht, pp. 433–491.
EIA, Energy Information Administration—U.S. Department of Energy, 2008. U.S.
Carbon Dioxide Emissions from Energy Sources. /http://www.eia.doe.gov/
oiaf/1605/flash/flash.htmlS.
EPA, U.S. Environmental Protection Agency, 2006. Greenhouse Gas Emissions from
the U.S. Transportation Sector: 1990–2003. /http://www.epa.gov/oms/climate/
420r06003.pdfS.
Ferretti, I., Zanoni, S., Zavanella, L., Diana, A., 2007. Greening the aluminum supply
chain. International Journal of Production Economics 108 (1–2), 236–245.
Figliozzi, M.A., 2011. The impacts of congestion on time-definitive urban freight
distribution networks CO
2
emission levels: results from a case study in
Portland, Oregon. Transportations Research–Part C 19 (5), 766–778.
Hall, R.W., 1987. Consolidation strategy: inventory, vehicles and terminals. Journal
of Business Logistics 8 (2), 57–73.
Hekkert, M.P., Joosten, L.A.J., Worrell, E., 2000. Reduction of CO
2
emissions by
improved management and product use: the case of transport packaging.
Resources Conservation and Recycling 30 (1), 1–27.
Higginson, J.K., Bookbinder, J.H., 1994. Policy recommendations for a shipment
consolidation program. Journal of Business Logistics 15 (1), 87–112.
Hoffman, W., 2007. Who’s carbon-free? Traffic World, 15 October 22.
Kohn, C., Brodin, M.H., 2008. Centralised distribution systems and the environment:
how increased transport work can decrease the environmental impact of logistics.
International Journal of Logistics: Research and Applications 11 (3), 229–245.
Linton, J.D., Klassen, R., Jayaraman, V., 2007. Sustainable supply chains: an
introduction. Journal of Operations Management 25 (6), 1075–1082.
Masters, J.M., 1980. Effects of freight consolidation on customer service. Journal of
Business Logistics 2 (1), 55–74.
McKinnon, A.C., 2000. Sustainable distribution: opportunities to improve vehicle
loading. UNEP Industry and Development 23 (4), 26–30.
McKinnon, A.C., 2003. Logistics and the environment. In: Hensher, D.A., Button, K.J.
(Eds.), Handbook of Transport and the Environment, Elsevier, Amsterdam,
pp. 665–686.
McKinnon, A.C., 2005. The economic and environmental benefits of increasing
maximum truck weight: the British experience. Transportation Research-D 10
(1), 77–95.
Merrick, R.J., Bookbinder, J.H., 2010. Environmental assessment of shipment
release policies. International Journal of Physical Distribution and Logistics
Management 40 (10), 748–762.
Murphy, P.R., Poist, R.F., 2003. Green perspectives and practices: a ‘‘comparative
logistics’’ study. Supply Chain Management: An International Journal 8 (2),
122–131.
Quariguasi Frota Neo, J., Bloemhof-Ruwaart, J.M., van Nunen, J.A.E.E., van Heck, E.,
2008. Designing and evaluating sustainable logistics networks. International
Journal of Production Economics 111 (2), 195–208.
RITA, 2008. Research and Innovative Technology Administration, Bureau of
Transportation Statistics. http://www.bts.gov/publications/transportation_
statistics_annual_report/2008/.
Samuelsson, A., Tilanus, B., 1997. A framework efficiency model for goods
transportation, with an application to regional less-than-truckload distribu-
tion. Transport Logistics 1 (2), 139–151.
Sarkis, J., 2006. Greening the Supply Chain. Springer-Verlag, NY, USA.
Sarkis, J., Zhu, Q., Lai, K-h., 2011. An organizational theoretic review of green
supply chain management literature. International Journal of Production
Economics 130 (1), 1–15.
Sbihi, A., Eglese, R.W., 2007. Combinatorial optimization and green logistics. 4OR:
Quarterly Journal of Operational Research 5 (2), 99–116.
Sheu, J.-B., Chou, Y.-H., Hu, C.-C., 2005. An integrated logistics operational model
for green-supply chain management. Transportation Research - Part E 41 (4),
287–313.
Srivastava, S.K., 2007. Green supply-chain management: a state-of-the-art litera-
ture review. International Journal of Management Reviews 9 (11), 53–80.
Sundarakani, B., de Souza, R., Goh, M., Wagner, S.M., Manikandan, S., 2010.
Modeling carbon footprints across the supply chain. International Journal of
Production Economics 128 (1), 43–50.
The Royal Society of London and the U.S. National Academy of Sciences, 1992. The
royal society and the national academy of sciences on population growth and
sustainability. Population and Development Review 18 (2), 375–378.
TRL, 1999. Methodology for Calculating Transport Emissions and Energy Con-
sumption. In: A.J. Hinckman (Ed.), Transport Research Laboratory, Crowthorne,
United Kingdom. Project Report SE/491/98. Deliverable 22 for the Project
MEET (Methodologies for Estimating Air Pollutant Emissions from Transport).
Ubeda, S., Arcelus, F.J., Faulin, J., 2010. Green logistics at Eroski: a case. Interna-
tional Journal of Production Economics. doi:10.1016/j.ijpe.2010.04.041.
¨
Ulk ¨
u, M.A., 2009. Comparison of typical shipment consolidation programs:
structural results. Management Science and Engineering 3 (4), 27–33.
¨
Ulk ¨
u, M.A., 2010. Analysis of Shipment Consolidation: Cost-Saving Opportunities
in Logistics Supply Chains. Lambert Academic Publishing, Germany.
¨
Ulk ¨
u, M.A., Bookbinder, J.H., Modeling shipment consolidation and pricing deci-
sions for a manufacturer–distributor, International Journal of Revenue Man-
agement, forthcoming.
Van Hoek, R., 1999. From reversed logistics to green supply chains. Supply Chain
Management: An International Journal 4 (3), 129–134.
Wu, H.-J., Dunn, S.C., 1995. Environmentally responsible logistics systems. Inter-
national Journal of Physical Distribution and Logistics Management 25 (2),
20–38.
Yang, B., Yang, Y., Wijngaard, J., 2005. Impact of postponement on transport: an
environmental perspective. The International Journal of Logistics Management
16 (2), 192–204.
M.A. ¨
Ulk¨
u / Int. J. Production Economics 139 (2012) 438–446446
