Copyright ⓒ2016 Korean Society of Civil Engineers
DOI 10.1007/s12205-016-2560-4 pISSN 1226-7988, eISSN 1976-3808
www.springer.com/12205
Potential CO
2Savings by Increasing Truck Size: A Korean Case Study
Nam Seok Kim
*, Bart Wiegmans
**, and Lei Bu
***Received July 27, 2015/Accepted Janaury 7, 2016
···
Abstract
In this paper, we present a method to estimate CO2 reduction potential by increasing the truck size. Trucks are apparently the most
preferred freight transport option for most shippers. Therefore, increasing the truck size may be a realizable and practical strategy, except when logistics companies (truck owners) tailor their truck sizes to customer needs despite their inefficiencies. However, increasing the truck size is not justified in all situations. Some types and sizes of trucks may fit specific distance ranges. The distance range in which a certain type and size of truck shows the highest efficiency can be determined by the break-even distance of the corresponding truck type/size. Using this information on break-even distances, the government can roughly estimate the potential CO2 savings. Based on a case study of only pallet shipping trucks, if a subsidy of a 10% discount of new truck purchase costs is given
to 1,000 truck owners for 10 years to foster an increase in the sizes of their trucks, the net amount of CO2 emissions that can be saved
by 2020 would be 103,069 t. Even though the quantity is not significant, the government expects that shippers and truck owners will more rationally select truck sizes if the information of the break-even distances is provided to the trucking market.
Keywords: truck, break-even distance, CO2, logistics, freight
···
1. Introduction
The Korean government has considered two key strategies to reduce CO2 emissions in the freight transportation sector by
2020. The first is the modal shift. It has been widely accepted that the modal shift is one of the most effective and sustainable ways to reduce CO2 emissions in the freight sector (Mckinnon,
2003). It has been proven that CO2 emissions per tkm between
trucks and rail- and vessel-based intermodal systems are considerable (Mckinnon, 2003; Van Essen et al., 2003; McKinnon and Piecyk, 2010; Park et al., 2012; Kim and Van Wee, 2014). However, when the size of the country is insufficient to producing economies of distance for rail or inland waterways, such as Korea, a modal shift to the rail/inland waterway does not occur. This is due to time and cost disadvantages, such as high drayage costs and delays at transshipment points (Kim and Van Wee, 2011). As a secondary strategy, a more realistic approach would be to increase the efficiency of trucks rather than changing the major transport mode. This is because trucks continue to be the most preferred freight transport option for most shippers (Teo, Taniguchi, and Qureshi, 2014). The most common way to increase the efficiency of trucks is to increase their sizes, despite the relatively lower impact on CO2 savings that this would have compared to
the modal shift. It is evident that larger trucks are more
energy-efficient than smaller trucks if they are fully loaded (Mckinnon, 2003; Odhams et al., 2010; Woodrooffe et al., 2010; EC, 1999; ATRI, 2011).
This study was motivated by the following research questions: How much CO2 can be saved if a specific percentage of trucks
are more efficiently operated by increasing their sizes? What types/sizes of trucks should be increased and why? Is it realistic to aim to attain the target amount of CO2, which is approximately
0.5 M ton? To address these questions, we designed three research steps. First, the concept of the break-even distance is applied to the trucking field. Generally, the concept of the break-even distance has been applied as a comparison among different freight transportation modes or systems (Kim and Van Wee, 2011; Rutten, 1995). It is assumed that a specific type/size of truck is chosen in a specific distance range in terms of the minimum unit cost. This distance range is determined by the respective break-even distances. The break-even distances of all pre-classified trucks are estimated. In other words, it is known what types/sizes of trucks have the minimum unit costs in specific distance ranges. This can be considered a theoretic (ideal) economic distance range (for a type of truck), which can differ from the distance range observed in the real trucking market. Second, the actual distance traveled according to the types/sizes of the trucks was examined by using a trucking industry survey conducted by the
TECHNICAL NOTE
*Assistant Professor, Dept. of Transportation and Logistics Engineering, Hanyang University, Ansan 426-791, Korea (E-mail: nskim@hanyang.ac.kr) **Assistant Professor, Dept. of Transport Planning, Faculty of Civil Engineering and Geosciences, TU Delft, The Netherlands (Corresponding Author, E-mail:
b.wiegmans@tudelft.nl)
***Assistant Professor, Dept. of Civil and Environmental Engineering, Jackson State University, Jackson 39217-0168, Mississippi, USA (E-mail: leibu04168@gmail.com)
Korea Transport Institute. This can be considered an actual distance range (for a type of truck). Finally, the potential CO2
savings in the trucking industry achieved by increasing the truck sizes is estimated by considering these two magnitudes (i.e., the differences between the total CO2 emissions from actual distances
and those from theoretical (ideal) distances).
2. Literature Review: Break-even Distance
2.1 Intermodal Freight System Break-even Distance Rutten (1995) defined the break-even point as the distance at which the costs of intermodal transport equal the costs of truck-only transport. The basic concept is quite simple, as shown in Eq. (1).
TC A = TC B (1)
where TC A is the total cost of A transport mode or system (e.g.,
truck only), and where TC B is the total cost of B transport mode or system (e.g., intermodal transport).
Figure 1 shows the cost competition and break-even distance between a rail-based (or barge-based) intermodal system and truck-only system. Owing to the two cost factors (i.e., drayage and transshipment costs of the intermodal system), calculating the intermodal break-even distance is more complicated than that of its counterpart.
Point D is the break-even point at which the total cost of an intermodal system is equal to the total cost of a truck-only system. In other words, before the break-even point, the truck-only system is economical, whereas after that point, the intermodal system is better. The lower slope of the rail or barge part indicated as d describes the economies of distance such that, in general, as the distance increases, the total cost decreases. The higher slopes denoted as aHO and aHD show the drayage rate, which is more expensive than the long-haulage truck rate. At B and C, the transshipment costs occur. Thus, the break-even distance is a result of the trade-off between the long-haulage trucking cost and intermodal cost. This is determined by the long-haulage trucking rate, b, and the drayage rates, aHO and aHD,
as well as by the transshipment costs, cHO and cHD, and rail or
barge rates.
2.2 Break-even Distance Comparison Among Trucks As in the intermodal and truck-only systems, the break-even distance of a given truck can be found when it is compared to other trucks. This conceptual approach is the same as the case of the intermodal system, except the former does not consider drayage and transshipment. The first break-even point between types A and B is indicated as A, as shown in Fig. 2. The total costs of the two truck types are the same in this case. In other words, before this break-even point, the type A truck (small truck) is economical. However, after this point (i.e., for a longer distance) the type B truck is more economical. The break-even distance indicated as B can be interpreted in the same way. This phenomenon can be explained with fixed and variable costs (Fig. 2). It is assumed that a variable cost is proportional to the distance (i.e., a constant rate).
3. Market Observation on Trucking Distance for
Analysis
3.1 Classification of Truck Types
One problem in applying this concept to the trucking market is that there are too many different types/sizes of trucks to be compared. Thus, trucks must be classified by type in a simpler way. To this end, the first standard is the Federal Highway Administration’s development of 13 classes in terms of the gross vehicle weight rating (GVWR) and numbers of axles, tires, and single and multiple units (Randall, 2012).
Secondly, according to the directives and regulations on motor vehicles by the European Commission (EC, 2014; UKdft, 2009), two types of trucks, categories N and O, exist. Category N is comprised of “motor vehicles with at least four wheels designed and constructed for the carriage of goods,” while category O is comprised of trailers (including semi-trailers). There are three category N subcategories: “Vehicles designed and constructed for the carriage of goods and having a maximum mass: not exceeding 3.5 t; exceeding 3.5 t but not exceeding 12 t; exceeding
Fig. 1. Cost Structure for Intermodal System and Truck-only Sys-tem (Kim and Van Wee, 2011)
Fig. 2. Concept of Break-even Distance in a Comparison of Truck Types
12 t.” There are four category O subcategories: “Trailers with a maximum mass: not exceeding 0.75 t; 0.75 t but not exceeding 3.5 t; 3.5 t but not exceeding 10 t; exceeding 10 t.” Vehicle classification by the Ministry of Land, Infrastructure, and Transport of the Republic of Korea, as shown in Fig. 3, outlines 12 vehicle classes, which is similar to the FHWA classification system (Randall, 2012; MLIT, 2013).
However, these classifications do not address the requirements of the present research topic. The criterion for classification in this study is the maximum payload (i.e., capacity in weight). Thus, trucks are reclassified by type and payload, as shown in Table 1.
The cargo truck type (Table 1: ID column A1–A9) has no enclosed body and is popular in Korea. Van and wing body trucks (indicated as T 1–5 and W 1–4) are panel trucks with fully enclosed bodies. They differ in terms of where the gate is installed for loading and unloading. The average box loaded in each truck except the cargo truck type is estimated the payload of the truck in m3 divided by average box size in m3 which is obtained through market observation (0.045 m3). The average pallets loaded in different size/type of trucks are obtained by observing trucking market (e.g. HOMEPLUS, a warehouse
distribution company invested by TESCO). For a tractor-trailer combination (C 1–12), C1 – 5 and C10-12 are the container-loadable trucks while C6 – 9 do not allow container. In previous studies (Odhams et al., 2010, Woodrooffe et al., 2010), W and T types are deemed rigid; C 1–6 and C9 are often referred to as a single trailer or workhorse; and C 7–8 and C 10–12 are referred to as double or high capacity. In addition, an issue with manufacturers exists. The five main manufacturers considered in the analysis are Benz, Daewoo, Hyndai, Scania, and Volvo. Their market share is greater than 95% of the Korean trucking industry. The averages of their different truck sale prices, fuel consumption rates, and sizes used in the analysis are summarized in Table 1. The information about how much it costs to operate different types of trucks is very important in the scope of this study and the truck sale prices crucially affect the total operation cost. It is required to look into the variability of the truck sales cost later. 3.2 Matching Commodity/unit Load with Truck Type/size Because the sizes and characteristics of commodity/unit loads vary, the commodity/unit load must be matched with the truck type/size. Fig. 4 depicts a visualization of unit loads and trucks (Table 1 outlines the truck IDs). When examining the trucking market’s focus on increasing the sizes of trucks, we determined that the truck shipment of containers does not compete with that of other unit loads, such as parcels and pallets. Thus, the break-even distances should be separately estimated.
The single commodity, along with the parcel, pallet, and container, are matched with several types of trucks. This should be specified because significant errors occur when the differences between the respective sizes of loading units and trucks are ignored. For example, we observed the Less Than Truckload (LTL) market and obtained the average magnitude of the number of parcels for several trucks. This is reported in Table 1 in the column of the average number of boxes loaded in a truck. Both pallets and containers are reported in the pallet and TEU columns. The fuel efficiency in the average fuel efficiency column is the average value of several truck manufacturers.
4. Methodology
4.1 Potential Number of Upsizing Trucks
In this section, we present a method of estimating the potential number of trucks that would be more beneficial if the size is increased based on the concept of the break-even distance. A simple example is shown in Fig. 5. There are three truck sizes: type A (small), type B (medium), and type C (large). There are three break-even points in which three segments are made: less than A; between A and B; and greater than B (the first, second, and third segment, respectively). In the first segment, the type A truck is most beneficial (i.e., incurs the lowest cost). In the second and third segments, types B and C are respectively the most efficient in terms of total cost. This basic concept can be applied for any combination of three or more consecutive trucks in the sizes shown in Table 1 (e.g., A6–A7–A8; W1–W2–W3–
Fig. 3. Vehicle Classification by the Korean Ministry of Land, Infra-Structure, and Transport (MLIT, 2013)
W4; C1–C2–···–C12).
Let, α2, β2, and γ2 be the numbers of type 2 trucks (medium
sized) observed when travelling “Before, Within, and After the corresponding break-even segment, respectively (vehicle). Additionally, let δ2, ε2, and ζ2 be loading units of α2, β2, and γ2,
respectively (either the numbers of parcels, numbers of pallets, or TEU). λi denoted the desired (optimized) i type of trucks which is fit to the range in terms of the break-even distances.
We observed the type 2 and determined that it is most efficient to be run in the second segment rather than in other segments because its unit cost is the lowest. However, some type 2 trucks would be observed either in the first or third segment in the real market. For the first segment, those trucks are certainly inefficiently operated in terms of the unit cost. Thus, the number of type 2 in the first segment is replaced with another truck type that would be the most efficient in the segment (in this case, a type 1). To convert a type 1 from a type 2, the number of loading units shipped in the 2 type is estimated (i.e., the number of type 2
Table 1. Reclassification of Trucks for Analysis Purposes (KW = Korean Won; 1 USD = 1,100 KW)
ID Vehicle type Length Width Height Pay-load (t) Pay-load (m3) Avg. box loaded* Pallet TEU Avg. fuel efficiency (km/liter) Total fixed cost (KW) Fixed cost per day (KW) Variable costs (KW/km)
A1 Cargo 1t 2.8 1.6 0.4 1 1.64 N/A N/A N/A 7.7 13,3225,000 9,184 251.3
A2 Cargo 1.4 t 3.1 1.6 0.4 1 1.79 N/A N/A N/A 8.3 0 0 234.9
A3 Cargo 2.5 t 4.3 1.9 0.4 3 3.14 N/A N/A N/A 6.4 35,130,000 18,297 304.7
A4 Cargo truck 3.5 t 4.9 2.1 0.4 4 3.80 N/A N/A N/A 5.5 36,460,000 16,880 352.2 A5 Cargo truck 5 t 4.6 2.3 0.4 5 4.20 N/A N/A N/A 4.4 58,303,333 24,293 440.5 A6 Cargo truck 8 t 7.3 2.3 1.4 8 23.91 N/A N/A N/A 4.5 75,050,000 28,428 433.3 A7 Cargo truck 11 t 9.1 2.3 1.4 11 29.81 N/A N/A N/A 3.0 103,066,667 35,787 643.6 A8 Cargo truck 18 t 10.1 2.3 1.4 18 33.09 N/A N/A N/A 2.7 122,700,000 39,327 720.9 A9 Cargo truck 25 t 10.1 2.3 1.4 25 33.09 N/A N/A N/A 2 140,766,667 45,118 975.0
T1 Van truck 1 t 3.1 1.8 1.7 1 9.00 201 N/A N/A 8.5 14,120,000 8,405 229.4
T2 Van truck 1 t enhanced 3.4 1.9 1.8 1 11.30 252 N/A N/A 8.2 15,240,000 9,071 237.8 T3 Van truck 1.2 t 3.8 1.8 2.0 1 13.04 291 N/A N/A 7.9 18,000,000 10,714 246.8 T4 Van truck 2.5 t 4.3 1.9 1.9 3 15.73 351 N/A N/A 6.8 41,550,000 21,641 286.8
T5 Van truck 5t 6.2 2.3 2.3 5 32.09 716 N/A N/A 4.9 59,570,000 27,579 398.0
W1 Wing body truck 2.5 t 4.2 1.8 1.8 3 13.61 304 N/A N/A 6.5 48,890,000 20,371 300.0 W2 Wing body truck 5 t 6.0 2.2 2.1 5 26.45 590 10 N/A 5.15 66,900,000 23,229 378.6 W3 Wing body truck 8t 8.0 2.3 2.1 8 39.31 878 14 N/A 4.05 84,983,333 29,508 481.5 W4 Wing body truck 11 t 9.1 2.3 2.1 11 44.72 998 18 N/A 3.1 115,300,000 40,035 629.0 C1 trailer – 20-ft. container 5.9 2.4 2.4 16 33.28 743 8 1 4 151,350,000 48,510 487.5 C2 Trailer – 40-ft. container 12.1 2.4 2.4 24 67.96 1517 20 2 3.5 155,850,000 49,952 557.1 C3 Trailer – 45-ft. container 13.6 2.4 2.4 25 76.48 1707 24 2.25 3.4 157,350,000 50,433 573.5 C4 Trailer – 40-ft. high cubic 12.1 2.4 2.7 25 76.46 1707 20 2 3.4 156,850,000 50,272 573.5 C5 Trailer – 45-ft. high cubic 13.6 2.4 2.7 25 86.04 1920 24 2.25 3.3 158,450,000 50,785 590.9 C6 Trailer - Double deck trailer 17.7 2.4 2.4 25 85.26 1903 30 N/A 3 230,000,000 73,718 650.0 C7 Multi-trailer1: 11 t + 5 t wing body 16.0 2.4 2.4 25 93.71 2092 28 N/A 3.4 230,000,000 73,718 573.5 C8 Multi-trailer 2: 8 t + 8 t wing body 16.0 2.4 2.4 25 102.91 2297 28 N/A 3.4 230,000,000 73,718 573.5 C9 Multi-trailer 3: Mega trans 13.3 2.5 2.6 26 84.07 1877 24 N/A 3 200,000,000 64,103 650.0 C10 Multi-trailer 2: 20 ft. + 20 ft. 11.8 2.4 2.4 32 66.39 1482 16 2 2.5 171,250,000 54,888 780.0 C11 Multi-trailer 3: 20 ft. +40 ft. 17.9 2.4 2.4 40 101.16 2258 28 3 2.3 174,450,000 55,913 847.8 C12 Multi-trailer 4: 20 ft. + 20 ft. + 20 ft. 17.7 2.4 7.2 48 99.59 2223 24 3 2 179,650,000 57,580 975.0 *Avg. box loaded = payload of the truck (m3)/one-box size (m3)
(Source: Manufacturers’ websites, Market observation)
trucks is converted into maximum loading units). Representing the symbols in Fig. 5, α2 is converted into δ2. The estimated
maximum loading unit (δ2, in this case) is reconverted into the
number of type 1 trucks which would be better option than type 2. As the symbols, δ2 is converted to λ1. In general, what is
important is to determine γi (we do not need to consider αi
because it is equal to γi-1). Note, there is some possibility that β2
would be different from λ2 since some β2 that might be inefficiently operated also need to convert to the other types of trucks.
4.2 CO2 Reduction Mechanism by Increasing the Truck
Size
It would be more efficient if types 1 and 3 are used in the first and third segments instead of a type 2 trucks. Specifically, if á2
and ã2 are replaced with α1 and γ3, respectively, it might be the
optimal case. Thus, the potential reduction in CO2 emissions is
the difference between the amount of CO2 emission produced by α2
and γ2 and the amount of CO2 emission produced by λ1 and λ3. The
units of α2, γ2, λ1 and λ3 are the numbers of vehicles that can be
converted to the number of trips. When the truck size increases, the number of trips is reduced. It is worthwhile to show a simple test that ACEA conducted (Larsson, 2009). The test primarily serves to determine how many trips are required for transporting 106 EU pallets (600 kg/pallet) according to different vehicle sizes. A 3.5-t truck in GCW must conduct 53 trips, whereas a 26-t truck in GCW requires only six trips. This result would be reduced to two trips if the truck of 60 in GCW is respectively used. This mechanism will work well for fixed routes between unchanged Origin–Destination (OD) pairs on a regular basis.
5. Analysis and Case Study
Information that is critical for estimating potential CO2
emissions is the updated numbers of trucks; e.g., λ1 and λ3 in the
example above. The criteria to find the updated numbers of trucks are the break-even distances. One may attempt to estimate all the break-even distances for all types of trucks. However, shifting from one type of truck to another occurs in very limited loading units. In this study, three types of loading units are considered, as shown in Fig. 5: parcels, pallets, and containers. Thus, it makes sense that the break-even distances are separately estimated from among the groups considered. Each group is associated with a type of network; i.e., a group of (1) parcel-pickup and delivery trucks; (2) pallet shipping trucks; and (3) container shipping trucks.
For all later analyses, some assumptions are commonly applied. First, the costs of the trucks are based on market observation, such as average fuel efficiency (km/liter), total fixed cost (Korean Won; KW (USD = 1,100 KW in January 2015), fixed cost per day (KW), and variable cost (KW/km) in the year of 2012, as shown in Table 1. Second, the price of diesel in 2012 was 1,950 KW/liter (about 2 USD/liter), which was expensive compared to the price in the year in which this research was conducted.
5.1 Break-even Distance Estimation for Three Groups 5.1.1 Break-even Distances of Parcel Pickup and Delivery
Trucks (group 1)
The groups of trucks in which parcels can be shipped are T1– 4, A1–9, and W1–2. Of this wide range, only T1–3 were compared because they were the most popular truck group in Korea according to pickup and delivery market observation. We assumed that the pickup and delivery distance was less than 70 km and that one box size is 0.045 CBM. Additionally, because the quantity to be shipped is another variable that influences both the choice of truck size and its break-even distance(Abate and de Jong, 2014), four demand scenarios of 10,000, 20,000, 50,000, and 100,000 parcels were designed (Fig. 6).
The break-even distance of T2 (a van type with a 1-ton payload and larger space compared to T1) was determined to be approximately 80 km in the case of 10,000 boxes (Fig. 6(a)), while it decreased to approximately 40 km when the box demand was doubled (Fig. 6(b)). As the number of boxes increased to 50,000 and 100,000, the break-even distance of T3 was approximately 80 km and 48 km, respectively. T2 showed the lowest unit cost between 15 km and 80 km for 50,000 boxes, and between 9 km and 46 km for 100,000 boxes. It is clear that the break-even points of trucks with lower payloads were lower. In other words, T2’s break-even points were 80 km, 40 km, 15 km, and 9 km as demand increased from 20,000 boxes to 100,000 boxes. 5.1.2 Break-even Distances of Pallet Shipping Trucks
(group 2)
The groups of trucks that accommodated load pallets were W2–4 and C1–12. Although pallets are possibly loaded in C types, the most frequently observed type and size of truck group
Fig. 5 Concept of Break-even Distance in a Comparison of Truck Types
in short and medium distances were W types. Of the C type, C1– 9 were added for examining higher demand levels and long-distance effects. Thus, W2 to 4 and C1 to 9 were included in the analysis, as shown in Fig. 4.
Because demand remains the most sensitive factor, two scenarios were examined: the number of pallet shipping of 2,500 and 8,000 per year. The former was approximately 10 pallets per
day, as shown in Figs. 7(a) and (b), while the latter was approximately 30 pallets per day, as shown in (c) and (d). No competition was observed for the short distance; therefore, the distances in all the graphs within Fig. 7 are presented from 150 km. Figs. 7(b) and (d) are enlarged versions of (a) and (c), respectively. In Fig. 7(b), W2 is the most efficient type/size before 175 km, while W3 showed the lowest unit shipping cost
Fig. 6. Break-even Distances of Pickup and Delivery Trucks (group 1): (a) Pickup and Delivery (# of Box = 10,000), (b) Pickup and Deliv-ery (# of Box = 20,000), (c) Pickup and Delivery (# of Box = 50,000), (d) Pickup and Delivery (# of Box = 100,000)
Fig. 7. Break-even Distances of Pallet Shipping Trucks (group 2): (a) Pallets (# of Pallet = 2,500), (b) Pallets (# of Pallet = 2,500), (c) Pal-lets (# of Pallet = 8,000), (d) PalPal-lets (# of Pallet = 8,000)
between 175 km and 192 km. After 192 km, C3 was the most efficient mode. C8 in (b) shows a very low slope. In other words, C8 could be a better option if either the number of pallets or the distance increased. This was confirmed, as shown in Fig. 7(d). The break-even distance of C8 was approximately 205 km, thereby surpassing C3 as the demand increased to 8,000 pallets. 5.1.3 Break-even Distances of Container Shipping Trucks
(group 3)
Containers can be shipped on only C types categorized in Table 1. Among the 12 options, only 4 types were considered: C1, C2, C10, and C11. C3 to C9 are made for pallet shipping purposes rather than container shipping. C12 was tested; however, it showed almost similar results as C11. Therefore, C12 was removed in the graph. Except for C1, the three modes competed in the 100-km range. C11 showed the lowest unit cost after 70 km. In the case of 200 TEU (not presented here), C11 also showed the lowest unit cost after approximately 40 km. Therefore, regardless of demand, multi-trailers are better than other single tractor-trailer systems.
5.2 Data of Actual Distances Travelled by All Truck Types In the previous sections, three reduction projects based on loading units and shipping characteristics (i.e., three groups) were analyzed. The key issue is to find γi (we do not need to
consider αi because it is equal to γi-1). Freight Transport Market Research Center in The Korea Transport Institute (2015) examines the average shipping distances for all types/sizes of trucks with sampling every year. Thus, the percentage of specific truck types that exceed the corresponding break-even distances is readily found. However, it does not make sense that all trucks outside the break-even range will naturally shift to the ‘fit’ size of the truck. The shifted quantity should be discussed with the government intervention described in next chapter.
6. Discussions
6.1 Policy Implementation
Although a type of mode (in this study, a type/size of truck) may be evidently superior to its competitors, the government cannot force truck owners or shippers to change the size of the
Fig. 8. Break-even Distances of Container Shipping Trucks (group 3): (a) Containers (700 TEU), (b) Containers (700 TEU)
Table 2. Strategy for Potential CO2 Reduction for Group 2: Pallet Shipping Trucks
Year Fuel consumption per truck (liter) 2012 2013 2014 2015 2016
Maximum number of trucks subsidized 1,000 1,000 1,000 1,000 1,000
Type/size shift Before After 22,576 500 500 500 500 500 W2, W3 W4 W4 C1, C2, C3 29,148 200 200 200 200 200 C1, C2, C3 C81 41,987 300 300 300 300 300 Potential CO2 (t) 11,452 11,452 11,452 11,452 11,452 Accumulated potential CO2 (t) 11,452 22,904 34,356 45,808 57,261
Year Fuel consumption
per truck (liter) 2017 2018 2019 2020
Maximum number of trucks subsidized 1,000 1,000 1,000 1,000
Type/size shift Before After 22,576 500 500 500 500 W2, W3 W4 W4 C1, C2, C3 29,148 200 200 200 200 C1, C2, C3 C8 41,987 300 300 300 300 Potential CO2 (t) 11,452 11,452 11,452 12,599 Accumulated potential CO2 (t) 68,713 80,165 91,617 103,069
1An increase to other multi-trailers is also beneficial. However, it is too complicated to separately estimate it. Because the difference in CO
2 is not too
trucks. In many cases, government indirectly supports truck drivers by providing subsidies. Specifically, the government can identify some truck owners who inefficiently operate their trucks according to the break-even distances. Secondly, the government can provide subsidies for the identified truck owners (not for all, but for some volunteers based on a first-in first-served basis) to increase the size of their trucks. The potential number of size-increased trucks, as shown in Table 2—for example, group 2—is used to estimate the number of trucks (e.g., γi) as receivers of the
subsidies. The potential CO2 reduction according to a plan is also
shown in Table 2.
We previously analyzed and discussed estimation of the amount of CO2 emissions that can be potentially saved through truck size
changes. One analytical assumption adopted in this study is the concept of the break-even distance. In general, the strategy to increase the truck size is to decrease the break-even distance. We suggest a strategy with two variants to shorten the break-even distance: 1) policies to reduce fixed costs (indicated as ‘F’ in Fig. 9, later on ‘F policy’), and 2) policies to reduce variable costs (indicated as ‘V’ in Fig. 9, later on ‘V policy’).
An example of ‘F policy’ is to provide a purchase subsidy for a specific type of truck that predominantly fits within a certain distance range (i.e., discounting a specific percentage of truck purchase costs results in a fixed cost discount). ‘V policy’ can be used to impose a so-called ‘a km-based CO2 tax’ or an increase
in fuel price (i.e., a variable cost change). Specifically, imposing a tax or increasing the fuel price simultaneously decreases the variable cost of type A trucks. However, the overall effect should lead to a greater decreasing slope for B types rather than for A types. Therefore, to reduce CO2 in logistics, the strategy of
increasing the truck size can be considered as giving a purchase subsidy, which is the ‘F policy’. Unlike a direct policy, such as those of ‘F’ and ‘V policy’ some indirect policies exist that are hardly categorized as having a fixed/variable cost: a change in permissible payload (Abate and de Jong, 2014), technology subsidies, and inexpensive financing (Guerrero et al., 2013).
In this study, we assume that the government will decide to provide subsidies to a very small number of truck owners for a limited shipping group (1,000 truck owners per year for group
2). As a result, at least in a numerical example, the amount of reduction until 2020 is expected to be approximately 103,069 t. The accumulated number of subsidy receivers would be 10,000, which is less than 3% of the entire registered truck owners in Korea. The potential amount of CO2 is estimated as 0.1 M,
which is smaller than the target amount of CO2 (0.467 M t).
The questions initially asked were the following: How much CO2 can be saved if a specific percentage of trucks are more
efficiently operated by increasing their sizes? What types/sizes of trucks should be increased? Is it realistic to aim for the target amount of CO2, which is approximately 0.5 M t? The answer to
the last question seems optimistic. By slightly increasing the subsidy receiver group, the target would be realistic. However, the strategy to shift to larger trucks through a ‘subsidy’ does not seem desirable in the long term. It would be more beneficial for truck owners if the government provided a guideline based on the estimation of break-even distances of several types/sizes of trucks and expected fuel savings according to the estimation. Truck owners could decide to adjust their truck sizes according to their major/regular O-D distances if they fully know the break-even distance information. Consequently, a better effect could be expected, although it is beyond the scope of this study. Nevertheless, it is important that this project is undertaken on a voluntary basis. Truck owners and shippers always pursue better efficiency. This means that the seemingly inefficient operation with smaller trucks could be customized and optimized for some truck owners/shippers. Therefore, it would be risky to assume that truck operations outside the break-even range are always inefficient. In addition, it is notable that the role of the government is still debatable. It was demonstrated throughout Figs. 6, 7, and 8 that truck carriers might optimally choose bigger trucks voluntarily whether there is a subsidy or not. In a perfect competitive market that all the information is well shared among truckers and shippers, the size of trucks is adjusted according to the distance and frequency of trips (orders). However, in the unstable trucking industry like one in Korea, such shifting does not happen well. The subsidy is a sort of catalyst or stimulus to the market. Some might argue there are much smoother options to lead the trucking market at equilibrium where trucks are optimally allocated because the subsidy policy for encourage to adjust the size of trucks is not directly target CO2 reduction.
6.2 Other Overlooked Issues: Trade-off
This study overlooked some important issues. The first issue is to compare the break-even distance based on internal (operational cost) only with the other break-even distance based on the full (social) cost including other external cost such as greenhouse gases, air pollutants (NOx, PM, NOx), noise, traffic accident and so on. The second issue is the trade-off between some benefit gained from bigger trucks and some extra cost to road pavement damage which should not be ignorable. Taking into account infrastructure damage, LCA (life cycle assessment) could be a future study.
7. Conclusions
In this paper, we presented a model for estimating the amount of potential CO2 emissions by increasing the size of trucks. It is
known that larger trucks are more energy efficient than smaller trucks if they are fully loaded (Mckinnon, 2003; Odhams et al., 2010; Woodrooffe et al., 2010; EC, 1999; ATRI, 2011) therefore, the government expects to reduce a specific amount of CO2 if
they guide truck owners who seem to inefficiently operate their trucks. One way to guide truck owners is to provide them with a subsidy when they replace their old trucks for new ones. The concept of the break-even distance is used as the criteria for subsidy distribution.
In addressing the initial questions, it seems possible to meet the target based on the assumption that 1,000 truck owners will receive a new truck purchase subsidy (i.e., a 10% discount in this study); nevertheless, many uncertainties remain. One scenario is that the government may decide to expand or shrink the size of the subsidy; the other is that some truck owners might voluntarily shift their trucks toward greater efficiency once the break-even information is uncovered.
Acknowledgements
This work was supported by the research fund of Hanyang University (HY-201400000002079-N).
References
Abate, M. and de Jong, G. (2014). “The optimal shipment size and truck size choice – The allocation of trucks across hauls.” Transportation Research Part A: Policy and Practice, Vol 59, pp. 262-277, DOI: 10.1016/j.tra.2013.11.008.
ATRI (2011). National petroleum council future transportation fuels study: truck demand group findings.
CE Delft (2003). To shift or not to shift that's the question, CE Delft, 03.4360.09.
EC (1999). Methodology for calculating transport emissions and energy consumption - MEET, Project Report SE/491/98.
EC (2014). Directives and regulations - motor vehicles <http://ec.europa. eu/ enterprise/sectors/automotive/documents/directives/motor-vehicles/ index_en.htm>.
FHWA (2012). Traffic recorder instruction manual: A, vehicle classification using FHWA 13-Category Scheme.
Freight Transport Market Research Center in The Korea Transport Institute. http://roft.koti.re.kr/ (Accessed Jan, 2015).
Guerrero, S. E., Madanat, S. M., and Leachman, R. C. (2013). “The trucking sector optimization model: A tool for predicting carrier and shipper responses to policies aiming to reduce GHG emissions.” Transportation Research Part E: Logistics and Transportation Review, Vol. 59, pp. 85-107, DOI: 10.1016/j.tre.2013.08.005. Kim, N. S. and Van Wee, B. (2011). “The relative importance of factors
that influence the break-even distance of intermodal freight transport systems.” Journal of Transport Geography, Vol. 19, No. 4, pp. 859-875, DOI: 10.1016/j.jtrangeo.2010.11.001.
Kim, N. S. and Van Wee, B. (2014). “Toward a better methodology for assessing CO2 Emissions for intermodal and truck-only freight systems: A european case study.” International Journal of Sustainable Transportation, Vol. 8, No. 3, pp. 177-201, DOI: 10.1080/15568318. 2011.633689.
Larsson, S. (2009). Weight and dimensions of heavy commercial vehicles as established by Directive 96/53/EC and the European Modular System (EMS), ACEA (European Automobile Manufactures Associaion). Logistics Research Centre (2010). Measuring and managing CO2
emissions.
Mckinnon, A. C. (2003). Logistics and the Environment, in Hensher, D. A. & Button, K. J. (Eds.) Handbook of Transport and the Environment. Elsevier Ltd.
MLIT (Ministry of Land, Infrastructure, and Transport) (2013). Manual of data collection for road traffic <http://www.law.go.kr/admRulLsInfoP. do?admRulSeq=2000000023225#AJAX>.
Odhams, A. M. C., Roebuck, R. L., Lee, Y. J., Hunt, S. W., and Cebon, D. (2010). “Factors influencing the energy consumption of road freight transport.” Proceedings of the Institution of Mechanical Engineers, Part C: Journal of Mechanical Engineering Science, Vol. 224, No. 9, pp. 1995-2010.
Park, D., Kim, N.S., Park, H., and Kim, K. (2012). “Estimating trade-off among logistics cost, CO2 and time: A case study of container transportation systems in Korea.” Journal International Journal of Urban Sciences, Vol. 16, No. 1, pp. 85-98, DOI: 10.1080/12265934. 2012.668322.
Rutten, B. J. C. M. (1995). On medium distance intermodal rail transport, Delft, Delft University of Technology.
Teo, J. S. E., Taniguchi, E., and Qureshi, A.G. (2014). “Multi-agent systems modelling approach to evaluate urban motorways for city logistics.” International Journal of Urban Sciences, Vol. 18, No. 2, pp. 154-165, DOI: 10.1080/12265934.2014.929020.
UKdft (2009). Definition of vehicle categories: Extracted from 2007/46/ EC as last amended by 385/2009 http://www.dft.gov.uk/vca/ vehicletype/definition-of-vehicle-categories.asp.
Woodrooffe, J., Glaeser, K.-P., and Nordengen, P. (2010). “Truck Productivity, Efficiency, Energy Use, and Carbon Dioxide Output: Benchmarking of International Performance.” Transportation Research Record: Journal of the Transportation Research Board, Vol. 2162, DOI: 10.3141/2162-08.
