Accepted Manuscript

Impact of fuel price and emissions on inventory policies

Amulya Gurtu, Mohamad Y. Jaber, Cory Searcy

PII: S0307-904X(14)00391-6
DOI: http://dx.doi.org/10.1016/j.apm.2014.08.001
Reference: APM 10105

To appear in: Appl. Math. Modelling

Received Date: 9 February 2014

Revised Date: 23 July 2014
Accepted Date: 13 August 2014

Please cite this article as: A. Gurtu, M.Y. Jaber, C. Searcy, Impact of fuel price and emissions on inventory policies,
Appl. Math. Modelling (2014), doi: http://dx.doi.org/10.1016/j.apm.2014.08.001

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
 Impact of fuel price and emissions on inventory policies

1 Impact of fuel price and emissions on inventory policies

2
3
4 Amulya Gurtu, Mohamad Y. Jaber, Cory Searcy
5
6 Department of Mechanical and Industrial Engineering, Ryerson University, 350 Victoria Str.,
7 Toronto, ON, M5B 2K3, CANADA
8
9 Abstract: The purpose of this paper is to analyze the impact of changes in fuel prices
10 and the imposition of a carbon tax on emissions from transport on shipment lot sizes
11 and supply chain costs. An analysis is done to show that increases in fuel prices should
12 be dealt with differently than other costs. Further, a function to calculate future fuel
13 prices has been developed. This function has been used to calculate transport cost in
14 the future. The EOQ models have been modified to include increasing transport cost
15 and emission tax to demonstrate its impact on various inventory policies. Due to
16 increases in fuel prices, the cost of every subsequent order will also increase, thus
17 resulting in an increase of average order cost for all the shipments in a production
18 cycle. Organizations that have their vendors in relatively close proximity will be at an
19 advantageous position in managing their supply chain costs more effectively in the
20 future. On the other hand, organizations that have invested heavily in global supply
21 chains will need to re-examine their supply chain strategy to overcome cost
22 challenges. This research presents a new challenge for supply chains/logistics
23 management strategies for organizations with global supply chains.
24
25 Article Classification: Research paper
26
27 Keywords: Order cost, Global supply chain, Economic Order Quantity (EOQ), Inventory
28 policy, Fossil fuel, Fuel price.
29
30
31
32 1. Introduction
33 There is a growing emphasis on the search for substitutes of fossil fuel that generate less
34 pollution, are available in abundant quantities and as efficient or more efficient as fossil fuel.
35 Nowadays, newspapers in many countries are full of news and editorials about the possible
36 depletion of the world’s fossil fuel reservoirs and the future of fuel prices. Many of the available
37 inventory management models, such as the economic order quantity (EOQ), developed over the
38 past century were based on the false assumptions that fossil fuel is abundant and that greenhouse
39 gas (GHG) emissions from manufacturing and logistics operations have no implicit effects [1].
40 Literature on EOQ, its application under varying degrees of shelf-life, yield and its limitations
Page 1 of 34
 Impact of fuel price and emissions on inventory policies

1 has been reviewed by many authors [2, 3, 4, and 5]. Despite many limitations, the EOQ is still in
2 use and has been celebrated recently in a book and in a special issue [6, 7, and 8].
3
4 One of the limitations of the EOQ model that has not been addressed in the literature is how an
5 increase in fossil fuel prices will affect transportation cost and lot sizes. Increasing fuel prices are
6 putting pressure on logistics and supply chain costs. Fuel cost has emerged as one of the top ten
7 challenges in the truck transport industry in the USA [9], accounting for 35% of the total
8 transportation cost in 2013[10]. Andriolo et al. [11] reached a similar conclusion that transport
9 cost is increasing and becoming an important factor in lot size and inventory decisions; they
10 suggested including it in inventory decisions. The greater the distance travelled, the higher the
11 cost of transportation and the bigger the impact on the environment. Oglethorpe and Heron [12]
12 reached a similar conclusion for national food chains. They also noted that “miles per unit of
13 consumption” is generally ignored, which is an important measure for reducing emissions.
14
15 The long term effects of increasing fuel prices on the total cost of supply chains have not been
16 addressed in sufficient detail. When a tax on transport emissions is considered, the total cost of
17 supply chains will increase further. A literature review on supply chains and logistics found little
18 evidence of theory development on the “total cost” of logistics. A survey article reviewing 683
19 supply chains and logistics papers found only a small fraction, 1.1%, focusing on the total cost
20 theory [13].
21
22 This paper attempts to fill this gap in the peer reviewed literature and provides an impact of
23 increases in fuel prices on transport cost, total order cost, and lot sizes. It further addresses the
24 impact of a tax on GHG emitted from transportation on the total order cost. The EOQ model has
25 been used to demonstrate the impact of increasing fuel prices and an emissions tax on the total
26 cost. The remainder of the paper is organized as follows: Section 2 reviews the literature, briefly
27 discusses how the demand and supply of fossil fuel affect its price, and subsequently the cost of
28 transporting goods. Section 2 also briefly discusses transport emissions and explains the reason
29 for treating fuel prices differently than the rest of the costs in supply chains. Section 3 discusses a
30 model for calculating the economic order quantity (EOQ) taking into account the changes in fuel

Page 2 of 34
 Impact of fuel price and emissions on inventory policies

1 prices as a function of time and provides a solution procedure. This model is then tested in a two
2 level-supply chain context with different coordination mechanisms. Section 4 provides
3 illustrative numerical examples that demonstrate the application of the model. Section 5 presents
4 analysis and discussion. Section 6 discusses managerial insights. Section 7 concludes the paper
5 with a discussion of the implications of the model on the future of supply chains.
6
7 2. Literature Review
8 In the context of a supply chain, transportation involves moving raw materials, semi-finished and
9 finished products from one entity (an origin) to another (a destination). The primary mode of
10 transportation for global supply chains is cargo shipping. Around 90% of the world trade is
11 carried by the international shipping industry [14]. However, emissions generated from
12 international aviation and maritime transports are ignored from national emissions inventories
13 [15]. The price of fuel influences the cost of cargo shipping as well as the configuration of
14 supply chains. The cost of bunker fuel in cargo shipping is about 30%, though this varies slightly
15 based on the speed and the size of the vessel, of the total cost of transportation, which is a
16 significant contributor to the overall cost [16]. Notteboom and Vernimmen [16] discussed the
17 effects of high fuel prices on marine cargo liners between the Far East and Europe and found that
18 an increase in fuel prices increases their operating cost and erodes their profitability. In a related
19 example, Jeihani and Sibdari [17] studied the impact of gas prices on the auto industry and found
20 that as the price of fuel increases, buyers switch to smaller cars to save on operating (fuel) cost.
21 Some researchers have investigated the effects of transportation costs on lot size decisions. For
22 example, Goyal and Szendrovits [18] modeled lot sizes based on the constraint of truck/container
23 capacity and fixed transport cost. They found that a firm could experience cost savings by
24 allowing both equal and unequal lot sizes at different stages of a production line (or a supply
25 chain) as compared to models that require equal lot sizes alone. Swenseth and Godfrey [19]
26 discussed inventory replenishment decisions due to price discounts on higher weight in
27 transportation cost and found that transporting in larger quantities saves money. Russell and
28 Krajewski [20] discussed the transport cost in a supply chain when calculating the lot size
29 quantity based on weight discounts. In the case of partial truckloads, they found that declaring
30 higher weights than the actual weight, cost less due to differences in rates. Ertogral et al. [21]

Page 3 of 34
 Impact of fuel price and emissions on inventory policies

1 added a step-function of the transport cost in a two-level (supplier-buyer) supply chain cost
2 function based on volume discounts with and without over-declaration. They found a potential
3 for savings by explicitly integrating the transportation cost into the total supply chain cost as it
4 alters inventory (and production) decisions. However, these studies, as others in the literature,
5 did not factor the changes in the price of fuel over time into their transportation cost equations.
6
7 Corbey and Jansen [22] studied the impact of various costs on the economic lot size quantity and
8 emphasized the inclusion of variable costs in the calculation. Zhao et al. [23] have worked on
9 adding transport costs whenever multiple trips per day, within cities or in close proximities, are
10 possible. They found adding fuel cost changes the lot sizes. Most of the supply chain literature
11 adopts a centralized policy approach by determining the joint sizing policy or the joint economic
12 lot-size problem (JELSP) that minimizes the total supply chain cost. Existing literature either
13 ignores direct fuel cost in the cost analysis or implicitly considers it a part of the fixed order
14 costs. This indicates a need for assessing the impact of changes in fuel prices, which affect the
15 cost of transportation, on the EOQ model. In support of this notion, readers may refer to the work
16 of Jaber and Zolfaghari [24] and Glock [25] for reviews on the JELSP.
17
18 Transportation is a major source of GHG emissions. Offshore outsourcing and global supply
19 chains have helped fuel the growth in emissions due to an increase in transportation needs. Davis
20 et al. [26] observed that 23% of the global CO2 emissions are embedded in traded goods and
21 suggested a CO2 emissions tax along the supply chain for reducing emissions. Chaabane et al.
22 [27] studied production processes and transportation systems in the aluminum industry and found
23 that the current legislation and Emissions Trading Schemes (ETS) need to be strengthened and
24 harmonized at the global level in order to drive a meaningful environmental strategy. Jaber et al.
25 [28] worked on the emissions in supply chains and presented a model for optimizing the cost of
26 emissions under various scenarios. The importance of designing more environmentally
27 responsible inventory models has been discussed by Bonney [29]. Later Bonney and Jaber [1]
28 also discussed various challenges in including the cost of impacts on the environment in
29 inventory management; they stressed the need for of assigning a cost of environmental impact
30 for taking meaningful decisions. The importance of the relationship between energy and

Page 4 of 34
 Impact of fuel price and emissions on inventory policies

1 inventory was discussed by Zanoni and Zavanella [30]. Benjaafar et al. [31] provided a generic
2 model to account for emissions associated with ordering, inventory holding, and production.
3 Chen et al. [32] further extended the work of optimizing the order quantities (EOQ) while
4 minimizing emissions. Hua et al. [33] studied the impact of carbon trading schemes on lot size.
5 To address a key gap in the literature, this paper assigns a cost to emissions from transportation
6 through a carbon tax and studies its impact, in combination with the variable fuel cost in
7 transport, on supply chain costs.
8
9 Ulku [34] acknowledged the need for improving environmental performance of transportation
10 and developed a model for consolidating the consignments, which showed improvements both in
11 cost and in environmental performance. Further, Pan et al. [35] stressed the need for reducing
12 GHG emissions from freight transport and developed a model for the pooling of supply chains as
13 one of the viable options for reducing global GHG emissions. Using data from two French retail
14 chains, Pan et al. [35] found a reduction in emissions by combining the two supply chains
15 through road and rail transport. Wahab et al. [36] addressed the environmental impact of
16 imperfect quality from an offshore vendor and suggested a dual supply chain (one offshore and
17 another onshore) as a solution. Rosic and Jammernegg [37] used a dual sourcing model (one
18 offshore and another onshore) for minimizing the environmental impact of transport. They found
19 that less sourcing from offshore improves the environmental performance of an organization,
20 though its economic performance is negatively affected. Piecyk and McKinnon [38] explored the
21 impact of road transport in supply chains on carbon emissions by 2020. They concluded that
22 some of the established practices, such as the centralization of manufacturing and warehousing,
23 JIT replenishment, and the outsourcing of non-core activities would not be sustainable. Bonney
24 and Jaber [1] also acknowledged the finite nature of fossil fuel and made similar comments about
25 these practices due to fuel/energy cost in the future. The European Commission provided a
26 detailed methodology for calculating emissions from transport [39]. However, Ubeda et al. [40]
27 acknowledged that it is difficult to precisely estimate fuel consumption and emissions from
28 transport and suggested the use of averages. They studied the impact of routing in relatively
29 close proximities for Eroski, a Spanish retailer, in reducing emissions. This paper uses average

Page 5 of 34
 Impact of fuel price and emissions on inventory policies

1 fuel consumption and tests the sensitivities for variation in average fuel consumption and travel
2 distances.
3
4 Fossil fuel, comprising of coal, oil, petroleum, and natural gas products, is still the single largest
5 source of energy, satisfying 87% for global needs [41]. A lot of progress has been made in the
6 area of renewable energy in the last decade; however, its use is mainly limited to stationary
7 applications and its share in overall energy consumption is negligible [41], 1.3% excluding hydro
8 and 7.8% including hydro in 2010. Fuel cell technology, which has the potential to replace fossil
9 fuel in transport in the future, is still in its infancy. This technology breaks water into oxygen and
10 hydrogen, and uses hydrogen as fuel. The handling of hydrogen in transport vehicles is a safety
11 concern and a challenge for designers. Although engineers are continuously striving for
12 improving the energy efficiencies of transport vehicles, improvements made in the efficient use
13 of fuel are offset by significantly higher demand. Figure 1 provides a summary of the trend in
14 global oil consumption which has been increasing steadily.
15
16 <Insert Figure 1 about here>
17
18 The possibility of exhaustion of oil and the associated concept of “peak oil” was given by
19 Hubbert [42, 43]. This is commonly referred to as “Hubbert’s Peak”, which is the point in time
20 where the maximum rate of global petroleum extraction is reached, after which the rate of
21 production enters terminal decline. The rate of global oil production has not increased in line
22 with the demand as shown in the Figure 2. Oil production in the last decade has been almost
23 constant while the demand has been increasing. This trend started well before the recent global
24 financial crisis, which started in the second half of 2008. This trend has put pressure on oil prices
25 and contributed to an increase in oil prices greater than overall inflation.
26
27 <Insert Figure 2 about here>
28
29 The price of a commodity is a function of demand and supply. Fuel prices are experiencing
30 pressure from both ends, i.e., increasing demands and depleting supplies. The increase in demand

Page 6 of 34
 Impact of fuel price and emissions on inventory policies

1 for energy due to industrialization in developing countries, the global increase in population and
2 the depletion of oil reserves are putting pressure on the prices of fossil fuel. Oil prices have been
3 steadily increasing over the years as shown in Figure 3. The best-fit curve for the price of oil per
4 barrel in US dollars is a function of the form y = 2.8917x + 11.802 where x is the number of
5 years counting from 1990.
6
7 <Insert Figure 3 about here>
8
9 <Insert Figure 4 about here>
10
11 Oil prices based on annual inflation, starting at every decade since 1950 until 2000, are given in
12 Figure 4. The oil price calculated based on annual inflation, with an exception of 1980, should
13 have been in the range of $21 to $34 per barrel in 2012. However, the annual average price of oil
14 in 2012 was about $86 per barrel. This indicates that oil prices are increasing at about 3 times the
15 rate of overall annual inflation (which includes oil). Therefore, fuel prices are required to be
16 considered differently for transport cost in supply chains. The price calculated based on annual
17 inflation from 1980 is an anomaly due to political disturbance in Iran in the late 1970s.
18
19 The cost of production as well as the corresponding cost of transportation will go up due to
20 increases in energy prices. Renewable sources of energy, such as solar, wind, tidal, and
21 geothermal, may dampen the rate of increase in the price of energy for industrial and domestic
22 purposes in stationary applications. However, the use of energy from any other source for
23 transport application in global supply chains is less likely to be available in the near term. Thus,
24 organizations that have outsourced their manufacturing activities to offshore vendors will most
25 likely face a challenge in determining workable inventory policies as fuel prices continue to
26 increase.
27
28 Organizations with offshore outsourced manufacturing and global supply chains have the
29 additional challenge of reducing GHG emissions from transport. One prominent illustration of
30 this challenge is that the European Union has proposed a regulation for reducing CO2 emissions

Page 7 of 34
 Impact of fuel price and emissions on inventory policies

1 from light commercial vehicles. The proposed regulation targets a CO2 emissions reduction to
2 175 g CO2/km by 2016 [45]. As a second example, the US Environment Protection Agency [46]
3 has new exhaust emissions standards (Subpart B, 1037.106) for CO2 for tractors above 26,000
4 pounds GVWR (gross vehicle weight rating). For a tractor with GVWR > 33,000 Pounds (16.5
5 Ton), high-roof sleeper cab, the maximum CO2 limit for years 2014-2016 is 75 g/Ton-mile CO2.
6 This emissions value is used in this paper to show its impact on the EOQ model.
7
8 3. The Model and Its Applications
9 In this section, the EOQ model is revisited and modified to account for fuel prices. Further, the
10 cost of emissions due to transportation has been added to the model. The applicability of the
11 model is investigated in a two-level (vendor-buyer) supply chain, where the sum of the players’
12 costs is jointly minimized by determining their respective lot size and shipping quantities. The
13 following notations have been used in the paper:
14
A0 Transport cost at the start of an order point $
A1 Annual incremental increase in transport cost $
Ai Transport cost for ith shipment $
Cm GHG Emissions for a transport carrier by a mode g/Ton/km
d Distance travelled (one way) Km
D Demand/consumption rate per unit time units/year
h Unit holding cost per unit of time; subscript b and v are added to $
show buyer and vendor
i Order number number
K Order cost per cycle; subscript b and v are added to show buyer $
and vendor
l Fuel consumption liters/km
n Vendor’s cycle number
P Production rate per unit time units/year
q Lot size units
*
q Optimal lot size with transport cost units
qE EOQ without transport and emissions costs units
qe Optimal lot size with transport cost and emissions cost units
Page 8 of 34
 Impact of fuel price and emissions on inventory policies

t Cycle time for a lot (fraction) [t=q/D] year

Tci Emissions tax rate $/Ton
th
TCi Cost of the i order with transport cost $
th
TCi(q) Total cost per unit time of the i order with transport cost $
th
TCie(q) Total cost per unit time of the i order with transport and $
emissions costs
α Fuel price at the time of the first order (from figure 3; α = $
(2.8917x + 11.802)/γ, where x is the time in years between the
first order since 1990)
β Annual incremental increase in fuel price (from figure 3; β = $
2.8917/γ,
γ Conversion factor from barrel to liter (1 oil barrel US = 159 liter) -
λ Buyer’s cycles in a single vendor’s (production) cycle, integer unit-less
1
2 3.1. The basic model with changing fuel price
3 As discussed in Section 2, the increase in fuel prices can no longer be ignored and will be
4 considered in inventory decisions. One common model in the existing literature is that the buyer
5 (level 1 in the supply chain) orders according to the EOQ policy. The basic model is developed
6 assuming that all other costs in manufacturing activities remains unchanged. This model is then
7 expanded for a two-level supply chain that consists of a vendor (or manufacturer) and a buyer in
8 the next section. The paper also modified the models of Banerjee [47] (referred onwards as B86),
9 Goyal [48] (referred onwards as G88), Hill [49] (referred onwards as H99), and Braglia and
10 Zavanella [50] (referred onwards as BZ03). B86 considered a lot-for-lot policy, G88 assumed
11 that a vendor’s cycle is an integer multiplier of the buyer’s order quantity and that no shipment
12 occurs during the production segment of a vendor’ cycle, H99 developed a model similar to that
13 of Goyal except that shipment occurs during the production segment of a vendor’ cycle, and
14 BZ03 applied a Vendor Managed Inventory (VMI) with consignment policy, where a lot is
15 shipped as soon as it is ready in a production cycle. These models have been selected because
16 they are extensions of Harris’ seminal work [51]. The concept of JELSP was initiated by Goyal
17 [52]. Hill [49] assumed that production batches would increase by a fixed factor while shipment
18 is possible both in equal and unequal lot sizes. Braglia and Zavanella’s [50] consignment stock

Page 9 of 34
 Impact of fuel price and emissions on inventory policies

1 (CS) model is a relatively newer concept where a buyer pays for what it has consumed. These are
2 used in industry where a vendor and a buyer have a long term relationship.
3
4 This paper assumes: (i) lead time is zero, (ii) infinite planning horizon, (iii) constant demand, (iv)
5 no shortages are allowed, (v) holding cost is unaffected by the change in oil prices, (vi) a lot is
6 delivered in a single shipment irrespective of truck/container capacity, (vii) transport cost is
7 separate from setup or order cost and is paid by the buyer, (viii) production rate is higher than the
8 demand rate.
9
10 Taking the fuel cost into account, the total cost for the ith order is the sum of the order cost,
11 inventory holding costs and transport cost, and is given by
hq
12 TCi = K + t + Ai (1)
2
13 where transport cost (Ai) can be written as
q
14 Ai = A0 + A1 (i − 1) (2)
D
15 where A0 = 2dlα, A1 = 2dlβ, 2 represents a round trip, d is the distance travelled in kilometers; l
16 is the fuel consumption in liters per km; α is the fuel price in $ at the time of the first order (from
17 figure 3), α = (2.8917x + 11.802)/γ with x being the time in years between the first order since
18 1990, γ is the conversion factor and is equal to 159 liters per barrel US, and β = 2.8917/γ is the
19 annual incremental increase in fuel price in $ (from figure 3). The total cost including transport
20 cost per unit time for a lot is obtained by dividing Eq. (1) by the cycle time, as
D hq D
21 TCi (q ) = K+ + Ai and replacing the value of Ai from Eq. (2) it can be written as
q 2 q
D hq D
22 TCi (q ) = K+ + A0 + A1 (i − 1) (3)
q 2 q
23 The optimal value of q that minimizes Eq. (3) is obtained by setting its first derivative with
24 respect to q equal to zero and solving for q to get

2 D ( K + A0 )
25 q* = (4)
h
26 Substituting the value of q* from Eq. (4) in Eq. (3) the minimum total cost will be
Page 10 of 34
 Impact of fuel price and emissions on inventory policies

1 TCi (q * ) = 2hD( K + A0 ) + A1 (i − 1) (5)

*
2 The EOQ formula without the transport cost is qE = 2 KD h , suggesting that q > q E .
3 Accounting for transportation cost recommends ordering in larger lots than suggested by the
4 EOQ formula. Furthermore, as previously suggested, transportation cost is a function of fuel
5 price, distance and fuel consumption, so the transport cost increases as the price of fuel, distance
6 or fuel consumption increase. Subsequently, the lot size increases. Generally, fuel consumption
7 per unit distance travelled, though marginally, reduces over time due to advancements in
8 technology but fuel price is a function of demand and supply, and distances in supply chains are
9 affected by globalization.
10
11 3.2. Emissions tax from transportation
12 Emissions are one of the consequences of fossil fuel-based transportation. Offshore outsourcing
13 further increases emissions due to the need for additional transportation. This section shows the
14 impact of adding an emissions tax, to neutralize the effects of emissions associated with
15 transportation, on lot sizes and supply chain costs.
16
17 It is assumed that emissions are not significantly affected by the load of the vehicle. The
18 emissions tax will be Te = 2dCmTci . The new cost function per unit of time for the ith order

D hq D D
19 TCi (q) = K + + Ai + Te (6)
q 2 q q
20 The lot size that minimizes Eq. (6) and the minimum total cost per unit time for the ith order are
21 given by
2 D ( K + A0 + Te )
22 q*e = (7)
h

23 TC i , e ( q * ) = 2 hD ( K + Te + A0 ) + A1 (i − 1) (8)

24 where subscript e stands for “with emissions”.

25

Page 11 of 34
 Impact of fuel price and emissions on inventory policies

1 3.3. Model for two-level coordinated supply chain

2 This section considers a two-level supply chain that consists of a vendor (manufacturer) and a
3 buyer. The players in the chain can either coordinate or choose not to coordinate their supply
4 chain. Since the non-coordinated supply chain results in higher cost, the model is developed for a
5 coordinated model known as the Joint Economic Lot Size (JELS) model [52]. The scenarios
6 explored and investigated are for a continuous increase in fuel price, which affects the order
7 costs (and therefore the unit price of an item). Banerjee [47] developed a cost function for finite
8 production rate and lot-for-lot production-shipment policy, i.e. i = n. Adding transport costs and
9 an emissions tax, the unit time cost function becomes:
D q D
10 ATCiB86 = ( Kb + Kv + Ai + Te ) +  hb + hv  (9a)
q 2 P
11 The optimal value of q that minimizes Eq. (9a) is obtained by setting its first derivative to zero
12 and solving for q to get
2D( Kb + K v + A0 + Te )
13 q B86 = (9b)
D
hb + hv
P
B86
14 Substituting the value of q from Eq. (9b) in Eq. (9a) to get

 D
15 ATCiB 86 = 2D( Kb + K v + A0 + Te ) hb + hv  + A1 (i − 1) (9c)
 P
16 Goyal [48] relaxed the assumption of lot-for-lot policy in Banerjee’s model [47], i.e. during the
17 production cycle a vendor makes λ shipments from a single setup. Graphical comparison of these
18 two models is given in Figure 5. Therefore, the cost per unit time for vendor setup and buyer
19 order will be ( K v + λKb ) D / λq . The vendor does not send the shipment to the buyer before the
20 inventory is finished because hb > hv . The buyer receives a shipment as soon as the inventory is
21 finished. The buyer’s average inventory remains at q/2.
22
23 <Insert Figure 5 about here>
24
25 The fuel price, and hence the transport cost, will increase between the first and subsequent
26 shipments from a single production lot. This will change the lot size from one shipment to the
Page 12 of 34
 Impact of fuel price and emissions on inventory policies

1 next shipment (within a production cycle). In order to avoid these increases between shipments
2 and keep the planning simpler in practice, the average fuel cost of all the shipments from a
3 production lot has been added as the variable transport cost for each shipment. Therefore,
4 transport costs from a single production lot will be the same for all shipments. Adding the
5 average transport cost and the emissions tax, Goyal’s [48] cost function, for the nth Vender’s
6 cycle, and lot size modifies to
D K  q  D  (2nλ − λ − 1)
7 ATCnG88 =  K b + v + A0 + Te  + (hb − hv ) + λhv 1 +  + A1 (10a)
q λ  2  P  2
8 It is clear that Eq. (10a) is convex where the partial second derivatives are positive for q > 0 and
9 λ > 0. The third term in Eq. (10a) is explained in Appendix A. The lot size quantity that
10 minimizes Eq. (10a) is obtained by setting the first derivative to zero and solving for q to get

 K 
2 D K b + v + A0 + Te 
11 q G 88 (λ ) =  λ  (10b)
  D 
(hb − hv ) + λhv 1 + 
  P 
G88
12 Replacing the value of q (λ) from Eq. (10b) into Eq. (10a), the cost function becomes

13  K   D 
ATC nG 88 (λ ) = 2 D  K b + v + A0 + Te  (hb − hv ) + λhv  1 +   + A1
(2nλ − λ − 1) (10c)
 λ   P  2

14 When λ=1, Eq. (10c), it reduces to ATCiB 86 making ATC iB 86 a special case of ATC nG 88 (where i =

15 n in B86). The optimal value of Eq. (10c) occurs when ATCnG88 (λ − 1) ≥ ATCnG88 (λ ) ≤ ATCnG88 (λ + 1)
16 .
17
18 Braglia and Zavanella [50] modeled an industrial strategy for inventory management using Hill’s
19 [49, 53] model, who suggested increasing lot size at a fixed rate for the optimal annual cost of
20 the supply chain. The behavior of inventory at the vendor’s and buyer’s sides for the models of
21 Hill [49] and Braglia and Zavanella [50] are shown in Figure 6. Braglia and Zavanella [50]
22 suggested a strategy of keeping inventories at the buyer’s end where it is managed by the vendor,
23 who is paid based on consumption (up to daily), which informs the vendor about the remaining
24 inventory and helps in planning the next production lot.
25
Page 13 of 34
 Impact of fuel price and emissions on inventory policies

1 <Insert Figure 6 about here>

2
3 Adding transport cost and the emissions tax in Hill’s model [49], the cost function and lot size
4 changes to:
D  Kv  D  P − D   (hb − hv )q (2nλ − λ − 1)
5 ATCnH 99 =  + Kb + A0 + Te  + qhv  + λ   + + A1 (11a)
q λ  P  2P   2 2

6 The lot size quantity that minimizes Eq. (11a) is given by setting the first derivative to zero and
7 solving for q to get

K 
D v + K b + A0 + Te 
8 q H 99 (λ ) =  λ  (11b)
 D  P − D  (hb − hv ) 
hv  + λ   + 
 P  2 P  2 

9 Substituting qH 99(λ) from Eq. (11b) in Eq. (11a), total cost function reduces to

K   D  P − D   (hb − hv )  (2nλ − λ − 1) (11c)

10 ATCnH 99 (λ ) = 2 D  v + K b + A0 + Te   hv  + λ   +  + A1
 λ   P  2P  2  2

11 The optimal value of Eq. (11c) occurs when ATCnH 99 (λ − 1) ≥ ATCnH 99 (λ ) ≤ ATCnH 99 (λ + 1) .
12
13 The expression for average transport cost from a production lot for Braglia and Zavanella’s
14 model [50] is different and explained in Appendix B. Adding the average transport cost function
15 to Braglia and Zavanella’s cost function, it becomes
D  Kv  D  P − D  (hb − hv )qi D
16 ATC nBZ 03 =  + K b + A0 + Te  + qi hb  + λ   −
q λ  P  2 P  2P

(λ − 1) D
17 + ( n − 1)λA1 + A1 (12a)
2 P

K 
D v + Kb + A0 + Te 
18 q BZ 03 (λ ) =  λ  (12b)
 D  P − D  (hb − hv )D 
hb  + λ   − 
 P  2 P  2P 

19
BZ 03
20 Substituting q (λ ) from Eq. (12b) in Eq. (12a), the cost function changes to
Page 14 of 34
 Impact of fuel price and emissions on inventory policies

K   D  P − D   (hb − hv )D 
2 ATC nBZ 03 ( λ ) = 2 D  v + K b + A0 + Te   hb  + λ   − 
 λ   P  2P  2P 

(λ − 1) D
3 + ( n − 1)λA1 + A1 (12c)
2 P
4 The optimal value of Eq. (12c) occurs when ATC nBZ 03 (λ − 1 ) ≥ ATC nBZ 03 (λ ) ≤ ATC nBZ 03 (λ + 1) . The
5 next section will provide the solution procedure.
6
7 3.4. Solution Procedure
8 This section provides a summary of the steps involved in finding the order cost and lot size after
9 adding the transport cost and the emissions tax.
10 Step 1: Define a starting date of the first production cycle, i.e., n=1.
11 Step 2: Calculate the values of α, β, A0, A1 and Te.
12 Step 3: Assume the value of λ = 1 as the starting point (because λ tends to become 1 as the
13 fuel price increases) to calculate total cost. Compute the corresponding ATCnM (λ )
14 for model M = G88, H99, and BZ03.
15 Step 4: Increase the value of λ in steps of 1. If the total cost ATCnM (1) < ATCnM (2) , select λ
16 = 1. Otherwise repeat the step for the next integer value of λ. Select the largest
17 value of λ for which ATCnM (λ ) ≤ ATCnM (λ + 1) where λ ≥1.
18 Step 5: Using the optimal value of λ from the previous step, calculate average shipment lot
19 size (q) for a production lot for the corresponding model M.
20
21 The next section presents numerical examples and the subsequent section discusses the results to
22 illustrate how these models behave with increasing fuel prices in order to draw some insights and
23 managerial implications.
24
25 4. Numerical Examples
26 The values of the input parameters have been taken from the numerical example used by Braglia
27 and Zavanella [50]. Then the variable transport cost and emissions tax have been added to
28 illustrate the impact of fuel price on order cost lot size. The notations used are:
Page 15 of 34
 Impact of fuel price and emissions on inventory policies

1
2 Demand rate (units/year) D 1,000
3 Production rate (units/year) P 3,200
4 Buyer’s order cost ($/order) Kb 25
5 Vendor’s setup cost for producing per run ($/batch) Kv 400
6 Holding cost for the vendor ($ /unit/ unit of time) hv 4.00
7 Holding cost for the buyer ($ /unit/ unit of time) hb 5.00
8
9 The emissions considered in the analysis are 75 g/Ton-mile CO2 for GVWR > 33,000 Pounds
10 (16.5 Ton) [45]. Therefore, emissions per kilometer for a 16.5 Ton trailer (1 mile = 1.6 km) =
11 75×16.5/(1.6×1000000) = 7.7344×10-4 Ton per km. Emissions tax at the rate of $30 per Ton of
12 CO2 equivalent has been used in the analysis [54]. Table 1 provides the values of order cost,
13 number of shipments per production cycle, shipment lot size with and without transport, and
14 emissions costs for ten production cycles. These costs are based on the assumption that the first
15 order is placed on Jan 1, 2014 for one-way travel distance (d) of 500 km taking fuel consumption
16 (l) as 0.3 Lt/km for the three policies discussed above. The data show that lot size more than
17 doubled for all the three policies. Order cost increases for every subsequent shipment due to
18 increase in fuel prices. In order to keep the cost of each shipment same, an average cost of
19 transportation for has been added to the function. This is done to avoid changes in the cost of
20 each shipment within a production lot where λ > 1. The details for finding the average cost of
21 transportation of all the shipments in a production cycle, i.e., where λ > 1 for Goyal’s and Hill’s
22 models, has been given in Appendix A, and for Braglia and Zavanella’s model in Appendix B.
23 Similarly, lot sizes in a cycle, where λ > 1, are the averages of all the shipments. The values for
24 Goyal’s model [48] and Braglia and Zavanella’s model [50], when λ=1, i.e. lot-for-lot
25 production, are the same as the values for Banerjee’s model [47]. The data show that Hill's
26 policy provided the least cost.
27
28 <Insert Table 1 about here>
29

Page 16 of 34
 Impact of fuel price and emissions on inventory policies

1 Subsequent results discussed in this paper refer to the average of ten production cycles, i.e.,
2 equivalent to the last line from the Table 1. Order costs and lot sizes increased as the distance
3 increased. Table 2 shows the changes in lot sizes and order costs by varying the distance. As the
4 distance increases, transport cost increases, which results in bigger and infrequent lot sizes.
5 However, approximately at 600 km, all the policies started behaving in the same manner, that is,
6 lot-for-lot. For the distance less than 600 km, Hill’s policy provided the least cost per unit. This
7 threshold of 600 km for first order placed in 2014 will decrease as the oil price goes up in
8 following years.
9
10 <Insert Table 2 about here>
11
12 Long haul transportation of goods is mostly done by trailers/containers. The fuel consumption
13 targets for a trailer up to 20 Ton and greater than 20 Ton are 3.09 km/L and 2.01 km/L,
14 respectively [45]. The modified models are tested for changes in (average) shipment lot sizes and
15 order cost by varying the fuel consumption in this range. The results are given in Table 3.
16
17 <Insert Table 3 about here>
18
19 5. Analysis and Discussion
20 The above analysis indicated that the impact of variable transport cost will become increasingly
21 important in supply chains. It also highlighted the need to consider GHG emissions, which are
22 rarely accounted for in existing models. By adding fuel and emissions costs, the shipment lot
23 sizes and order cost increased for the three tested models. This resulted in less frequent
24 shipments in larger sizes. As a consequence of the increase in order cost in each shipment, the
25 unit cost of a product will go up over time. The unit cost of a product will also increase as the
26 distance in a supply chain increases. The data from Table 3 indicates that the model is relatively
27 less sensitive to fuel consumption. An increase in fuel consumption by about 50% (for a fixed
28 distance) increased the annual order cost in the range of 4% to 7% for the three models.
29 However, the model is more sensitive to distance as depicted in Table 2. For a 40% increase in

Page 17 of 34
 Impact of fuel price and emissions on inventory policies

1 distance (from 500 to 700 km), the cost went up by 7% to 10% and for a further 43% increase in
2 distance (from 700 to 1000 km), the cost went up by 10% to 12% for the models.
3
4 There are a few options for reducing supply chain costs and improving environmental
5 performance. Reducing the distance between manufacturing and consumption locations is a
6 logical choice for reducing variable fuel cost as well as emissions. Nevertheless, it is not possible
7 to eliminate transportation from supply chains. It can only be reduced. Therefore, organizations
8 have to make an effort in developing a culture of being eco-oriented for reducing their carbon
9 footprint when designing and implementing logistics policies [55]. A carbon tax is an enabler in
10 improving the environmental performance. Choi [56] studied the impact of a carbon price on
11 sourcing in the fashion industry. The author found a positive influence of a carbon tax on
12 reducing environmental impact by sourcing from local manufacturers.
13
14 The next option is to invest in alternate and efficient means of transportation. In order to address
15 the need for reducing fuel costs, truck manufacturers have been developing more efficient
16 vehicles and exploring the use of alternate fuels in long haul trucks. For example, researchers
17 have recently developed a prototype car which can run up to 125 mph (200 kmph) using
18 chocolate as a fuel [57]. A Trans-Atlantic flight by a Swiss prototype solar powered plane is a
19 major breakthrough in harnessing solar energy for mobile applications and may have
20 considerable implications for transport vehicles in the future [58]. Use of Compressed Natural
21 Gas as an alternate fuel in long haul trucks is being explored for reducing emissions. Hybrid
22 electric trucks for city use have been developed [59]. Hybrid trucks for long haul applications are
23 an interesting potential solution to reduce emissions and consumption of fossil fuel. However,
24 hybrid trucks for long haul applications are still in various stages of development and have to
25 overcome many technical challenges of reliability, performance, safety, and cost in addition to a
26 psychological barrier of acceptance by the transport industry.
27
28 Another corollary to the above analysis is to use more efficient means of transport technology
29 that generates fewer emissions, such as railways. However, this requires a large investment in
30 infrastructure, which takes years to build and is feasible only between the locations of hubs of

Page 18 of 34
 Impact of fuel price and emissions on inventory policies

1 economic activities for inland operations only. A combination of hybrid vehicles and railways
2 would be a workable solution for surface transport. Cargo ships, which carry the biggest share of
3 goods in global supply chains, still remain a challenge in reducing costs and emissions. The next
4 sub-section provides a high level financial viability of a hybrid truck for a long-haul application.
5
6 6. Managerial Insights
7 The above discussion indicated that there are many practical uses of the findings of the impact of
8 increasing fuel prices and the possibility of a carbon tax on transport emissions in the real world.
9 One of the affects will be to evaluate the long term impacts of fuel prices on supply chain costs
10 and on sourcing strategies. In the situations where the cost an element is increasing rapidly or
11 volatility, vendor tries to charge at the highest anticipated rate and customer negotiates to pay on
12 an average price. It is a very likely scenario that a buyer would agree to pay either the fuel cost
13 on the prevailing price at the time of shipment or pay a surcharge, based on an index on fuel
14 price, at the time of shipment. Either method will treat fuel prices as variable. This will also
15 affect logistics/transport organizations in evaluating and selecting their transportation fleet. Long
16 haul hybrid trucks are in various stages of development and it is anticipated that the truck
17 industry will start switching to hybrid trucks gradually as the technology becomes reliable and
18 the cost of operating diesel trucks becomes economically unviable. A high level analysis on the
19 financial viability of a hybrid truck, based on the limited information publicly available from
20 Volvo on their long haul hybrid truck’s performance, is provided here. These are expected to
21 consume about 30% less fuel and, therefore, generate 30% less emissions [60]. A hybrid truck
22 with a comparable performance to a diesel truck will be an economically viable option in the
23 truck industry when the lifecycle cost, which includes acquisition, operations, maintenance and
24 salvage values, of a hybrid truck is less than the lifecycle cost of a diesel truck, i.e.,
25 C h + M h + Ah + E h − S h < C d + M d + Ad + E d − S d

26 Suffix d is for diesel trucks and h is for hybrid trucks. Cd and Ch are purchase prices, Md and Mh
27 are maintenance costs, Ad and Ah are fuel cost, Ed and Eh are emissions costs and Sd and Sh are
28 salvage value at the end of useful life of a diesel and a hybrid truck respectively. For the
29 purposes of comparison, the following values are for a Volvo diesel truck.
30

Page 19 of 34
 Impact of fuel price and emissions on inventory policies

1 Price of a diesel truck (Volvo 465 HP) $125,000

2 Average annual distance travelled (km/year) 200,000
3 Life of a truck (km) 2,000,000
4 Maintenance costs ($/year) 15,000
5 Fuel consumption (l/km) 0.30
6
7 The key assumption is that the annualized purchase price of a hybrid truck is double that of a
8 diesel truck due to high research costs, high warranty costs, and low initial volumes. Annual
9 maintenance costs are assumed to be 50% higher than that of a diesel truck. These costs are on
10 the higher side as a worst case scenario to avoid any surprises to transporters. Keeping the
11 emissions tax as discussed earlier, a diesel truck is economically a better choice than a hybrid
12 truck for about the next 5 years. The trend will start to shift in favour of hybrid truck beyond this
13 point, see Table 4. If the initial cost is reduced, fuel savings are greater, or the emissions tax rate
14 increases, the trend may shift in favour of a long haul hybrid truck sooner.
15
16 <Insert Table 4 about here>
17
18 Another possible impact, depending upon the pace of increase in fuel prices, is that globalization
19 of manufacturing and outsourcing activities may go in the reverse direction, i.e., fuel prices may
20 trigger localization. As an example, the increase in fuel prices will potentially severely affect the
21 cost of transportation of refrigerated items. This might increase the prices of perishable items,
22 such as meat and dairy products, significantly in the future. While these are examples of a few
23 possible negative effects, it should be acknowledged that there are possibilities of finding of an
24 alternate source of energy for mobile applications. Either way, the structure of supply chains and
25 logistics will be greatly affected by increases in fuel prices in the future.
26
27 7. Conclusions
28 From the foregoing discussion and the numerical example, it can be concluded that the order cost
29 and lot size are affected by changes in the fuel price, emissions tax, fuel consumption and
30 distances travelled. As the fuel price increases, the cost of every subsequent order will also

Page 20 of 34
 Impact of fuel price and emissions on inventory policies

1 increase resulting in an increase of the average order cost for all the shipments in a production
2 cycle. Fuel cost is proportional to the distance travelled and increases due to offshore
3 outsourcing. while fuel prices are increasing at a much higher rate. Therefore, supply chains with
4 large distances between the buyer’s and vendor’s locations will feel a large impact on their
5 supply chain costs. In addition to the fuel price, growing awareness on the environmental impact
6 of transport emissions might prompt governments to impose a higher tax on those emissions.
7 This insight has both strategic and operational implications for organizations. In order to
8 accommodate rapidly increasing oil prices, supply chain professionals will either adjust transport
9 cost at a much higher rate than overall inflation or pay for the fuel at the prevailing (higher)
10 prices at the time of shipment. Either way costs of transportation and supply chains will go up in
11 the short term, which is an operational issue. Organizations that have their vendors in relatively
12 close proximity will be at an advantageous position in managing their supply chain costs more
13 effectively in the future. On the other hand, organizations that have invested heavily in global
14 supply chains will need to re-examine their supply chain strategy to overcome cost challenges, as
15 increases in fuel prices impact their profitability. This may lead to changing vendor bases to a
16 nearby location or increasing the prices to customers, which will reduce their competitive
17 advantage or absorbing the increased cost of fuel, which will reduce their profitability.
18
19 Financial options and hedging are tactical tools for reducing volatility in fuel prices over a short
20 period. Technological development of transport vehicles/cargo ships, which generate fewer
21 emissions and use alternate fuels, thereby reducing dependence on fossil fuel, might represent a
22 long-term strategic solution. As discussed earlier, hybrid vehicles for long haul applications are a
23 strong contender for a technological breakthrough in surface transport and are likely to be
24 commercially available in the next 5 to 10 years. Technological changes will have significant
25 impact on the automotive, transport, logistics, and shipping industries and there will be many
26 challenges in supply chain management during this transition. However, acceptance of a new
27 technology, such as hybrid vehicles, is driven by economic viability rather than environmental
28 consciousness.
29

Page 21 of 34
 Impact of fuel price and emissions on inventory policies

1 As fuel prices increase, buyers will look for alternative vendors that are closer to the point of
2 consumption. This could fundamentally change the pace and direction of globalization. These
3 findings align with those of Bonney and Jaber [1], who note: “Perhaps the continued increase in
4 energy prices may result in deglobalisation of markets, if not decentralization. In other words, a
5 move towards smaller self-sustainable markets would be a strategic solution to consider." This
6 paper provides an understanding of the importance of fossil fuel prices in supply chain costs
7 which addresses a key gap in the literature. It also demonstrated the impact of increasing fuel
8 prices and an emissions tax on various inventory policies. Three polices were analyzed and the
9 results showed that the three policies converged to lot-for-lot policy.
10
11 The paper gave an example of truck transport. However, the model is adaptable to marine as well
12 as air transport in global supply chains. This paper could be extended in several ways. This paper
13 assumed that a shipment lot will go in one truck/container irrespective of the lot size. The paper
14 could be extended by relaxing this assumption and adding the constraint of a truck/container
15 capacity to the models presented. The paper can also be extended to investigate the effects of
16 emissions minimizing policies, such as a carbon tax or cap and trade scheme, on the changes in
17 transport technology. Other possible extensions include the estimation of time when corporations
18 will be forced to focus on regional markets due to increases in fuel and emissions costs; the
19 estimation of the lifecycle costs of a hybrid vehicle in long haul applications; emissions from the
20 use of hybrid vehicles in long haul transportation. Additional work could focus on using more
21 precise fuel consumption and emissions data rather than averages as done in this paper. A case
22 study using actual fuel consumption and emissions could potentially provide such data. A case
23 study would also provide an opportunity to investigate the model developed in this paper in a
24 real-world supply chain context. The model can also be extended to include other costs where
25 energy is used, such as in manufacturing and storage, particularly in refrigerated storage and/or
26 refrigerated transportation, where energy needs are significantly different.
27
28 Appendix A:
29 Considering a model with multiple shipments from a single production lot based on consumption
30 like Goyal [48] or Hill [49] (refer Appendix B for Braglia and Zavanella [50]) one has to take

Page 22 of 34
 Impact of fuel price and emissions on inventory policies

1 into consideration that the transport cost increases with every subsequent shipment to the buyer.
2 Transport cost to a buyer’s 1st shipment i=1, in the first vendor cycle, n = 1, would be
q q
3 Ai = A0 + A1 (i − 1) = A0 , the transport cost of the second shipment would be A0 + A1 , and
D D
q
4 the transport cost of the third shipment would be A0 + 2 A1 , etc. So, if the vendor sends λ
D
q
5 shipment to the buyer, then the cost of the λth shipment would be A0 + (λ − 1) A1 . In a vendor’s
D
λ
q λ −1 q
6 cycle, the total shipping cost is λA0 + ∑ (i − 1) A1 = λA0 + λ A1 . In the second vendor
i =1 D 2 D
2λ
q
7 cycle, n = 2, the total shipping cost is λA0 + ∑λ (i − 1)A
i = +1
1
D
=

2λ
q λ
q 2λ (2λ − 1) − λ(λ − 1) q
8 λA0 + ∑ (i − 1) A1 − ∑ (i − 1) A1 = λA0 + A1 , with the general
i =1 D i =1 D 2 D
9 expression for the total shipping cost in the vendor’s cycle n can be written as
nλ
q ( n −1) λ q q  nλ (nλ − 1) − λ ( n − 1)(λ ( n − 1) − 1)
10 λ A0 + ∑ (i − 1) A1 − ∑ (i − 1) A1 = λA0 + A1  
i =1 D i =1 D D 2 
q λ(2nλ − λ − 1)
11 = λA0 + A1
D 2
q (2nλ − λ − 1)
12 Therefore, the average transport cost for a shipment = A0 + A1
D 2
13
14
15 Appendix B:
16 In case of Braglia and Zavanella’s model [50], the shipments from a production lot take place
17 based on production of a shipment lot (q) and there is an interval between the last shipment from
18 a production lot and the first shipments from the following production lot. The method for
19 calculating transport cost for nth production cycles and ith shipment, where λ is the total
20 shipments from a production lot, is given below:
21
Production Cycle (n) Shipment (i) Transport Cost
1 1 A0

Page 23 of 34
Impact of fuel price and emissions on inventory policies

q
2 A0 + A1
P
q
3 A0 + 2 A1
P
q
4 A0 + 3 A1
P
.... ....
q
λ A0 + (λ − 1) A1
P
1 λ (λ − 1) q  (λ − 1) q
For n=1, average cost per shipment = λA0 + A1  = A0 + A1
λ 2 P 2 P
λq
2 1 A0 + A1
D
λq q
2 A0 + A1 + A1
D P
λq q
3 A0 + A1 + 2 A1
D P
λq q
4 A0 + A1 + 3 A1
D P
.... ....
λq q
λ A0 + A1 + (λ − 1) A1
D P
λq (λ − 1) q
For n=2, average cost per shipment = A0 + A1 + A1
D 2 P
λq
3 1 A0 + 2 A1
D
λq q
2 A0 + 2 A1 + A1
D P
λq q
3 A0 + 2 A1 + 2 A1
D P
λq q
4 A0 + 2 A1 + 3A1
D P
.... ....

Page 24 of 34
 Impact of fuel price and emissions on inventory policies

λq q
λ A0 + 2 A1 + (λ − 1) A1
D P
2λq (λ − 1) q
For n=3, average cost per shipment = A0 + A1 + A1
D 2 P
1
q (λ − 1) q
2 General expression for average cost per shipment = A0 + (n − 1)λA1 + A1
D 2 P
3

Page 25 of 34
 Impact of fuel price and emissions on inventory policies

1 References
2 [1] M. Bonney, M.Y. Jaber, Environmentally responsible inventory models: Non-classical
3 models for a non-classical era, International Journal of Production Economics, 133(1)
4 (2011) 43-53.

5 [2] M.Y. Jaber, Modeling Hidden Costs of Inventory Systems - A Thermodynamic Approach,
6 in: M.Y. Jaber (Ed.), Inventory management: Non-Classical Views, CRC Press, New York,
7 2009 , pp. 199–218.

8 [3] M.Y. Jaber, S. Zanoni, L.E. Zavanella, An entropic economic order quantity (EnEOQ) for
9 items with imperfect quality, Applied Mathematical Modelling, 37(6) (2013) 3982-3992.

10 [4] I. Konstantaras, K. Skouri, M.Y. Jaber, Inventory Models for Imperfect Quality Items with
11 Shortages and Learning in Inspection, Applied Mathematical Modelling, 36(11) (2012)
12 5334-5343.

13 [5] B. Maddah, K. M.K. Salameh, G.M. Karame. Lot Sizing with Random Yield and Different
14 Qualities, Applied Mathematical Modelling, 33(4) (2009) 1997-2009.

15 [6] T. Choi, Handbook of EOQ Inventory Problems - Stochastic and Deterministic Models and
16 Applications, Springer, New York, 2014.

17 [7] C.H. Glock, E.H. Grosse, J.Mm Ries, The lot sizing problem: A tertiary study,
18 International Journal of Production Economics, (doi:10.1016/j.ijpe.2013.12.009).

19 [8] A.A.Taleizadeh, M. Nematollahi, An Inventory Control Problem for Deteriorating Items

20 with Back-Ordering and Financial Considerations, Applied Mathematical Modelling, 38(1)
21 (2014) 93-109.

22 [9] ATRI, Critical issues in the trucking industry – 2013, Retrieved on Jan 07, 2014 from
23 http://atri-online.org/2014/01/28/csa-implications-2013-update-request/

24 [10] ATRI, ATRI- An Analysis of the Operational Costs of Trucking: A 2012 Update.
25 Retrieved on May 30, 2013 from http://atri-
26 online.org/?s=An+Analysis+of+the+Operational+Costs+of+Trucking%3A+2012+Update
27 &searchsubmit=

28 [11] A. Andriolo, D. Battini, A. Persona, F. Sgarbossa, R. W. Grubbström, A Century of

29 Evolution from Harris's Basic Lot Size Model: Survey and Research Agenda, International
30 Journal of Production Economics (doi: 10.1016/j.ijpe.2014.01.013).

31 [12] D. Oglethorpe, G. Heron, Sensible operational choices for the climate change agenda, The
32 International Journal of Logistics Management, 21(3) (2010) 538-557.

33 [13] C.C. Defee, B. Williams, W.S. Randall, An inventory of theory in logistics and SCM
34 research, The International Journal of Logistics Management, 21(3) (2010) 404-489.

Page 26 of 34
 Impact of fuel price and emissions on inventory policies

1 [14] IMO - Introduction to IMO. IMO. Retrieved on 27 Apr. 2013 from

2 http://www.imo.org/About/Pages/Default.aspx.

3 [15] UNFCCC. Issues - bunker fuels. UNFCCC. Retrieved 13 July 2013, from
4 http://unfccc.int/methods/emissions_from_intl_transport/items/1057.php

5 [16] T.E. Notteboom, B. Vernimmen, B, The effect of high fuel costs on liner service
6 configuration in container shipping, Journal of Transport Geography, 17(5) (2009) 325–
7 337.

8 [17] M. Jeihani, S. Sibdari, The impact of gas price trends on vehicle type choice, Journal of
9 Economics and Economic Education Research, 11(2) (2010) 1–11.

10 [18] S.K. Goyal, A.Z. Szendrovits, A constant lot size model with equal and unequal sized
11 batch shipments between production stages, Engineering Costs and Production Economics,
12 10(1) (1986) 203–210.

13 [19] S.R. Swenseth, M.R. Godfrey, Incorporating transportation costs into inventory
14 replenishment decisions, International Journal of Production Economics, 77(2) (2002)
15 113–130.

16 [20] R.M. Russell, L.J. Krajewski, Optimal purchase and transport cost lot sizing for a single
17 item Decision Science, 22(4) (1991) 940–952.

18 [21] K. Ertogral, M. Darwish, M. Ben-Daya, Production and shipment lot sizing in a vendor-
19 buyer supply china with transportation cost, European Journal of Operational Research,
20 176(3) (2007) 1592–1606.

21 [22] M. Corbey, E. Jansen, The economic lot size and relevant costs, International Journal of
22 Production Economics, 30(1) (1993) 519–530.

23 [23] Q. Zhao, S. Wnag, K.K. Lai, G. Xia, Model and algorithm of an inventory problem with
24 the consideration of transportation cost, Computer & Industrial Engineering, 46(2) (2004)
25 389–397.

26 [24] M.Y. Jaber, S. Zolfaghari, Quantitative Models for Centralised Supply Chain
27 Coordination, in: V. Kordic (Ed.), Supply Chain, Theory and Applications, InTech, Vienna,
28 Austria, 2008, pp. 307-338

29 [25] C.H. Glock, The joint economic lot size problem: A review, International Journal of
30 Production Economics, 135(2) (2012) 671–686.

31 [26] S.J. Davis, G.P. Peters, K. Caldeira, The supply chain of CO2 emissions, Proceedings of
32 the National Academy of Sciences of the United States of America, 108(45) (2011) 18554–
33 18559.

Page 27 of 34
 Impact of fuel price and emissions on inventory policies

1 [27] A. Chaabane, A. Ramudhin, M. Paquet, Design of sustainable supply chains under the
2 emission trading scheme, International Journal of Production Economics, 135(1) (2012)
3 37–49.

4 [28] M.Y. Jaber, C.H. Glock, A.M.A. El Saadany, Supply chain coordination with emissions
5 reduction incentives, International Journal of Production Research, 51(1) (2013) 69–82.

6 [29] M. Bonney, Inventory planning to help the environment, in: M.Y. Jaber (Ed.), Inventory
7 management: Non-Classical Views, CRC Press, New York, 2009, pp. 43–74.

8 [30] S. Zanoni, L. Zavanella, Energy and inventories, in: Jaber, M. Y. (Ed.), Inventory
9 management: Non-Classical Views, CRC Press (2009) pp. 43–74.

10 [31] S. Benjaafar, Y. Li, M. Daskin, Carbon Footprint and the Management of Supply Chains:
11 Insights From Simple Models, IEEE Transactions on Automation Science and
12 Engineering, 10(1) (2013) 99-116.

13 [32] X. Chen, S. Benjaafar, A. Elomri, The Carbon-Constrained EOQ, Operations Research

14 Letters, 41(2) (2013) 172–179.

15 [33] G. Hua, T. C. E. Cheng, S. Wang, Managing carbon footprints in inventory management,

16 International Journal of Production Economics, 132(2) (2011) 178–185.

17 [34] M.A. Ulku, Dare to care: Shipment consolidation reduces not only costs, but also
18 environmental damage, International Journal of Production Economics, 139(2) (2012)
19 438–446.

20 [35] S. Pan, E. Ballot, F. Fontane, The reduction of greenhouse gas emissions from freight
21 transport by pooling supply chains, International Journal of Production Economics, 143(1)
22 (2013) 86–94.

23 [36] M.I.M. Wahab, S.M.H Mamun, P. Ongkunaruk, EOQ models for a coordinated two-level
24 international supply chain considering imperfect items and environmental impact,
25 International Journal of Production Economics, 134(1) (2011) 151–158.

26 [37] H. Rosic, W. Jammernegg, The economic and environmental performance of dual

27 sourcing: A newsvendor approach, International Journal of Production Economics, 143(1)
28 (2013) 109–119.

29 [38] M.I. Piecyk, A.C. McKinnon, Forecasting the carbon footprint of road freight transport in
30 2020, International Journal of Production Economics, 128(1) (2010) 31–42.

31 [39] European Commission, Directorate-General Transport. Meet: Methodology for Calculating

32 Transport Emissions and Energy Consumption, 99 Vol., 1999.

33 [40] S. Ubeda, F.J. Arcelus, J. Faulin, Green logistics at Eroski: A case study, International
34 Journal of Production Economics, 131(1) (2011) 44-51.
Page 28 of 34
 Impact of fuel price and emissions on inventory policies

1 [41] BP-Statistical Review of World Energy 2012. BP. Retrieved on December 29, 2012 from
2 http://www.bp.com/assets/bp_internet/globalbp/globalbp_uk_english/reports_and_publicat
3 ions/statistical_energy_review_2011/STAGING/local_assets/pdf/statistical_review_of_wo
4 rld_energy_full_report_2012.pdf.

5 [42] M.K. Hubbert, Energy from fossil fuels, Science, 109(2823) (1949) 103–109.

6 [43] M.K. Hubbert, Nuclear Energy and the Fossil Fuels. Presented Before the Spring Meeting
7 of the Southern District, American Petroleum Institute, Plaza Hotel, San Antonio, Texas,
8 March 7-8-9, (1956).

9 [44] InflationData. Historical oil prices: InflationData.com Retrieved on 10 August 2013, from
10 http://inflationdata.com/Inflation/Inflation_Rate/Historical_Oil_Prices_Table.asp.

11 [45] IEA, Key World Energy Statistics-2010. IEA. Retrieved on 20 March 2012 from
12 http://www.tsl.uu.se/uhdsg/Data/IEA_Key_Stats/IEA2010.pdf

13 [46] ECFR — Electronic Code of Federal Regulations. ECFR — Code of Federal Regulations.
14 Retrieved on 23 Apr. 2013 from http://www.ecfr.gov/cgi-bin/text-
15 idx?c=ecfr&SID=7d5b4b07a299df765ea0353bea3a981f&rgn=div5&view=text&node=40:
16 34.0.1.1.4&idno=40.

17 [47] A. Banerjee, A joint economic-lot-size model for purchaser and vendor, Decision Sciences,
18 17(3) (1986) 292–311.

19 [48] S.K. Goyal, A joint economic-lot-size model for purchaser and vendor: a comment’,
20 Decision Sciences, 19(1) (1988) 236–241.

21 [49] R.M. Hill, The optimal production and shipment policy for the single-vendor single buyer
22 integrated production-inventory problem, International Journal of Production Research,
23 37(11) (1999) 2463–2475.

24 [50] M. Braglia, L. Zavanella, Modelling an industrial strategy for inventory management in

25 supply chains: The 'consignment stock' case, International Journal of Production
26 Research, 41(16) (2003) 3793–3808.

27 [51] F.W. Harris, How many parts to make at once, Factory, The Magazine of Management,
28 10(2) (1913) 135–136, & 152.

29 [52] S.K. Goyal, An integrated inventory model for a single supplier–single customer problem.
30 International Journal of Production Research, 15(1) (1977) 107–111.

31 [53] R.M. Hill, The single-vendor single-buyer integrated production-inventory model with a
32 generalised policy, European Journal of Operational Research, 97(3) (1997) 493–499.

33 [54] Carbon Tax, Province of British Columbia. Finance - Province of British Columbia.
34 Retrieved on 21 Nov. 2012 from http://www.fin.gov.bc.ca/tbs/tp/climate/A4.htm.
Page 29 of 34
 Impact of fuel price and emissions on inventory policies

1 [55] S.T. Kim, S.Y. Lee, Stakeholder pressure and the adoption of environmental logistics
2 practices, The International Journal of Logistics Management, 23(2) (2012) 238-258.

3 [56] T. Choi, Local Sourcing and Fashion Quick Response System: The Impacts of Carbon
4 Footprint Tax. Transportation research Part E, 55 (2013) 43-54.

5 [57] University of Warwick. Formula 3 Racing Car Powered by Chocolate and Steered by
6 Carrots. Retrieved on 16 May 2013 from
7 http://www2.warwick.ac.uk/newsandevents/pressreleases/racing_car.

8 [58] Solar impulse. Solar Impulse - around the world in a solar airplane. Retrieved on 21 May
9 2013. http://www.solarimpulse.com/en/.

10 [59] Iveco. Iveco Electric: New Daily Electric the future of urban transportation. Retrieved on
11 12 August 2013, from
12 http://web.iveco.com/uk/products/pages/daily_electric_features_and_benefits.aspx.

13 [60] Volvo. Volvo FE Hybrid. Retrieved on 21 March 2013 from

14 http://www.volvotrucks.com/trucks/global/en-gb/trucks/new-trucks/Pages/volvo-fe-
15 hybrid.aspx.

16

Page 30 of 34
 Impact of fuel price and emissions on inventory policies

1
2
3
4
5
6 Table 1. Changes in order cost and lot size due to transport and emissions cost.

Goyal (1988) Hill (1999) Braglia and Zavanella (2003)

Production lot
(n)
ATC nG 88 λn qG88 ATC nH 99 λn qH99 ATC nBZ 03 λn qBZ03

Without
transport 2,275 2 198 1,903 5 110 2,035 4 123
emissions cost
1 2,742 1 439 2,691 2 299 2,742 1 439
2 2,747 1 439 2,702 2 299 2,747 1 439
3 2,753 1 439 2,713 2 299 2,753 1 439
4 2,758 1 439 2,724 2 299 2,758 1 439
5 2,764 1 439 2,734 2 299 2,764 1 439
6 2,769 1 439 2,745 2 299 2,769 1 439
7 2,775 1 439 2,756 2 299 2,775 1 439
8 2,780 1 439 2,767 2 299 2,780 1 439
9 2,786 1 439 2,778 2 299 2,786 1 439
10 2,791 1 439 2,789 2 299 2,791 1 439
Average (1-10) 2,766 2,740 2,766
7
8

Page 31 of 34
 Impact of fuel price and emissions on inventory policies

1
2
3
4
5
6 Table 2. Changes in order cost and lot size due changes in transport distance in 2014.

Braglia and Zavanella

Goyal (1988) Hill (1999)
(2003)
Distance (km)
G 88 G 88 H 99 BZ 03 BZ 03
ATC avg λavg q avg H 99 λavg qavg ATC avg λavg q avg
ATC avg

0 2,275 2 198 1,903 5 110 2,035 4 123

100 2,404 1 384 2,149 3 182 2,256 2 232
200 2,499 1 398 2,327 2 256 2,413 2 247
300 2,591 1 412 2,472 2 271 2,562 2 261
400 2,680 1 426 2,609 2 285 2,680 1 426
500 2,766 1 439 2,740 2 299 2,766 1 439
600 2,851 1 451 2,848 1 409 2,851 1 451
700 2,933 1 464 2,933 1 464 2,933 1 464
800 3,013 1 476 3,013 1 476 3,013 1 476
900 3,091 1 487 3,091 1 487 3,091 1 487
1000 3,167 1 499 3,167 1 499 3,167 1 499
2000 3,858 1 602 3858 1 602 3,858 1 602
7
8

Page 32 of 34
 Impact of fuel price and emissions on inventory policies

1
2
3
4
5
6 Table 3. Effect of changes in fuel consumption on order cost and lot size.

Goyal (1988) Hill (1999) Braglia and Zavanella (2003)

Fuel
Consumption G 88 G 88 H 99 H 99 BZ 03 BZ 03
ATC avg λavg q avg ATC avg λavg qavg ATC avg λavg q avg
(Lt/km)

0.21 2,652 1 422 2,566 2 281 2,650 1 376

0.22 2,665 1 424 2,586 2 283 2,665 1 408
0.23 2,678 1 425 2,605 2 285 2,678 1 425
0.24 2,691 1 427 2,625 2 287 2,691 1 427
0.25 2,704 1 429 2,644 2 289 2,704 1 429
0.26 2,716 1 431 2,664 2 291 2,716 1 431
0.27 2,729 1 433 2,683 2 293 2,729 1 433
0.28 2,741 1 435 2,702 2 295 2,741 1 435
0.29 2,754 1 437 2,721 2 297 2,754 1 437
0.30 2,766 1 439 2,740 2 299 2,766 1 439
0.31 2,779 1 441 2,758 2 315 2,779 1 441
7
8

Page 33 of 34
 Impact of fuel price and emissions on inventory policies

1
2
3
4
5
6 Table 4. Comparison of annual costs of ownership for diesel and hybrid trucks.
Year Annual Cost of a diesel truck ($) Annual cost of a hybrid truck ($)
2013 111,709 116,643
2014 115,216 119,097
2015 118,936 121,701
2016 122,881 124,463
2017 127,066 127,392
2018 130,762 129,979
2019 135,470 133,275
2020 140,463 136,770
7

Page 34 of 34
Figure

Figures: Impact of fuel price and emissions on inventory policies

5000
4500 y = 44.73x + 2006.4
R² = 0.9031
4000
3500
Million Tons

3000
2500
2000
1500 World Oil Consumption

1000 Linear (World Oil Consumption)

500
0
1967
1969

2007
1965

1971
1973
1975
1977
1979
1981
1983
1985
1987
1989
1991
1993
1995
1997
1999
2001
2003
2005

2009
2011
2013
2015
2017
2019
2021
Figure 1. World oil consumption [41].

Page 1 of 6
Figures: Impact of fuel price and emissions on inventory policies

4,500
Flattened Production
4,000
3,500
3,000
Million Tons

2,500
2,000
1,500
1,000
500
0

2009
1965
1967
1969
1971
1973
1975
1977
1979
1981
1983
1985
1987
1989
1991
1993
1995
1997
1999
2001
2003
2005
2007

2011
Figure 2. World oil production [41].

Page 2 of 6
Figures: Impact of fuel price and emissions on inventory policies

120

100 y = 2.8917x + 11.802

Oil Prices (USD per barrel)

R² = 0.6473
80

60

40

Oil prices Linear (Oil prices)

20

Figure 3. Inflation adjusted oil prices [44].

Page 3 of 6
Figures: Impact of fuel price and emissions on inventory policies

100
90
Oil Price Per Barrel (USD)

86.5 (Nominal)
80 79.5 (1980)
70
60
50
40 34.0 (1990)
30 26.8 (1950)
23.7 (1960)
20 22.8 (2000)
10 20.8 (1970)
0
1988

2002
1950
1952
1954
1956
1958
1960
1962
1964
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1990
1992
1994
1996
1998
2000
2004
2006
2008
2010
2012
Nominal Price From 1950 From 1960
From 1970 From 1980 From 1990
From 2000

Figure 4. Oil prices based on annual inflation [44].

Page 4 of 6
Figures: Impact of fuel price and emissions on inventory policies

q Time

Figure 5. Inventory movement in Banerjee’s [47] and Goyal’s [48] models.

Page 5 of 6
Figures: Impact of fuel price and emissions on inventory policies

q
Time

Figure 6. Comparison of Hill’s [49] and Braglia & Zavanella’s [50] models.

Page 6 of 6