The Productivity Effects of Truck Size and Weight Policies

Click HERE for graphic.

ORNL- 6840

FINAL REPORT

THE PRODUCTIVITY EFFECTS OF TRUCK SIZE AND WEIGHT POLICIES

David P. Middendorf

Michael S. Bronzini

Center for Transportation Analysis

Energy Division

Oak Ridge National Laboratory

Oak Ridge, Tennessee 37831

November 1994

Managed by

Martin Marietta Energy Systems, Inc.

for the

Department of Energy

under Contract No. DE-AC05-850R21400

Prepared for

Federal Highway Administration

U.S. Department of Transportation

Washington, D.C. 2059

TABLE OF CONTENTS

Chapter Page

EXECUTIVE SUMMARY xv

1. INTRODUCTION 1

BACKGROUND 1

RESEARCH OBJECTIVES 1

ORGANIZATION OF THE REPORT 2

2. DATA COLLECTION AND LOGISTICS COST ANALYSIS METHODOLOGY 5

INTRODUCTION 5

SHIPPER SURVEY 5

Information Collected 5

Data Collection Methodology 7

Sample Size and Characteristics 10

FREIGHT TRANSPORTATION ANALYZER 13

3. EFFECTS OF LCV USAGE ON THE LOGISTICS COSTS OF TRUCK

SHIPPERS 19

INTRODUCTION 19

OVERALL EFFECTS 19

INFLUENCE OF VARIOUS COST, PRODUCT,AND LANE VARIABLES 22

Ratio of Freight to Inventory Costs 22

Product Value 25

Annual Lane Volume 26

Lane Distance 29

Annual Lane Ton-Mileage 29

Product Value and Annual Lane Volume 31

Product Value and Annual Lane Ton-Mileage 37

INTRAMODAL DIVERSION TO LCVs 38

4. EFFECTS OF LCV USAGE ON THE LOGISTICS COSTS OF RAIL AND

INTERMODAL SHIPPERS 45

INTRODUCTION 45

RAIL BOXCAR SHIPPERS 45

INTERMODAL SHIPPERS 51

5. ESTIMATION OF NATIONWIDE LCV USAGE 59

INTRODUCTION 59

INTRAMODAL DIVERSION MODEL 59

ESTIMATION PROCEDURE 60

iii

TABLE OF CONTENTS (Continued)

Chapter PAGE

5. ESTIMATION OF NATIONWIDE LCV USAGE (continued)

ESTIMATED INTRAMODAL DIVERSION 66

COMPARISON WITH OTHER ESTIMATES 71

6. CONCLUSIONS 75

REFERENCES 79

LIST OF FIGURES

Figure No. Page

1 Percent reduction in total logistics cost using 23

LCVs under low GVW limits as a function of freight

to inventory cost ratio for single trailers

2 Percent reduction in total logistics cost using 24

LCVs under high GVW limits as a function of freight

to inventory cost ratio for single trailers

3 Rail freight rate versus percent cost reduction 47

using Rocky Mountain doubles under existing

GVW limits

4. Rail freight rate versus percent cost reduction 48

using Rocky Mountain doubles under higher GVW limits

5. Rail freight rate versus percent cost reduction 49

using turnpike doubles under existing GVW limits

6. Rail freight rate versus percent cost reduction using 50

turnpike doubles under higher GVW limits

7. Intermodal freight rate versus percent cost reduction 54

using Rocky Mountain doubles under existing GVW limits

8. Intermodal freight rate versus percent cost reduction 55

using Rocky Mountain doubles under higher GVW limits

9. Intermodal freight rate versus percent cost reduction 56

using turnpike doubles under existing GVW limits

10. Intermodal freight rate versus percent cost 57

reduction using turnpike doubles under higher

GVW limits

LIST OF TABLES

Table No. Page

1. Distribution of traffic lane observations in the shipper 11

survey by type of product

2. Distribution of traffic lane observations in the shipper 12

survey by mode of transportation and type of equipment

3. Weight capacity of longer combination vehicles under two 14

gross vehicle weight (GVW) limit scenarios

4. Cubic capacity of longer combination vehicles compared 15

to single trailers of various sizes

5. LCV rate adjustment factors based on a comparison of 16

operating costs per loaded mile

6. Breakdown of FTA observations by current principal mode 17

7. Overall effect of LCVs on the total logistics cost of 20

single trailer truck shippers by type of LCV, GVW

limits, and relative LCV rates

8. Truck configuration resulting in the lowest total 21

logistics cost by GVW limits and relative LCV rates

9. Correlation between product value and the percent 26

reduction in total logistics cost resulting from

LCV usage

10. Correlation between percent cost reduction from using 27

LCVs and the composite variable formed by multiplying

product value by the inventory carrying cost expressed

as a percentage of inventory value

11. Average percent reduction in total logistics cost 28

from use of LCVs for different levels of annual lane

volume

12. Correlation between annual lane volume and the percent 29

reduction in total logistics cost resulting from LCV

usage

13. Average percent reduction in total logistics cost 30

from use of LCVs for different traffic lane distances

vii

LIST OF TABLES (continued)

Table No. PAGE

14. Average percent reduction in total logistics cost from 31

use of LCVs for different levels of annual lane

ton-mileage

15. Correlation between annual lane ton-mileage and the 32

percent reduction in total logistics cost resulting

from LCV usage

16. Combined effect of annual lane volume and product value 33

on total logistics cost using Rocky Mountain doubles

under existing GVW limits

17. Combined effect of annual lane volume and product value 34

on total logistics cost using Rocky Mountain doubles

under higher GVW limits

18. Combined effect of annual lane volume and product value 35

on total logistics cost using turnpike doubles under

existing GVW limits

19. Combined effect of annual lane volume and product value 36

on total logistics cost using turnpike doubles under

higher GVW limits

20. Combined effect of annual lane ton-mileage and product 39

value on total logistics cost using Rocky Mountain

doubles under existing GVW limits

21. Combined effect of annual lane ton-mileage and product 40

value on total logistics cost using Rocky Mountain

doubles under higher GVW limits

22. Combined effect of annual lane ton-mileage and product 41

value on total logistics cost using turnpike doubles

under existing GVW limits

23. Combined effect of annual lane ton-mileage and product 42

value on total logistics cost using turnpike doubles

under higher GVW limits

24. Percent of FTA cases and ton-mileage assumed to divert 43

to LCVs under different GVW limits, cost savings

thresholds, and relative LCV freight rates

25. Overall effect of LCVs on the total logistics cost of 46

rail boxcar shippers

26. Transportation mode with the lowest total logistics 46

cost for rail boxcar shippers

viii

LIST OF TABLES (continued)

Table No. Page

27. Correlation between rail freight charge per mile 51

and percent reduction in total logistics cost from

switching to LCVs

28. Overall effect of LCVs on the total logistics cost 52

of intermodal shippers

29. Transportation mode with the lowest total logistics 53

cost for intermodal shippers

30. Correlation between intermodal freight charge per 53

mile and percent reduction in total logistics cost

using LCVs

31. Percent of annual traffic lane ton-mileage that 61

would divert to LCVs

32. Definitions of relevant data items selected from 1987 63

TIUS public use records

33. Estimated truck vehicle-miles (in millions) diverting 67

to LCVs under existing GVW limits with LCV freight

rates same as current single trailer truckload rates

34. Estimated truck vehicle-miles (in millions) diverting 68

to LCVs under existing GVW limits with LCV freight

rates higher than current single trailer truckload rates

35. Estimated truck vehicle-miles (in millions) diverting 69

to LCVs under higher GVW limits with LCV freight

rates same as current single trailer truckload rates

36. Estimated truck vehicle-miles (in millions) diverting 70

to LCVs under higher GVW limits with LCV freight

rates higher than current single trailer truckload rates

37. Estimated billions of truck miles diverting to Turner 73

prototypes in the intercity dry van truckload

market section.

LIST OF ABBREVIATIONS AND SYMBOLS

AASHTO American Association of State Highway and

Transportation Officials

CFS Commodity Flow Survey

cwt hundredweight (hundreds of pounds)

EOQ Economic Order Quantity

FAK Freight All Kinds

FTA Freight Transportation Analyzer computer program

GVW Gross Vehicle Weight

LCV Longer Combination Vehicle

LTL Less Than Truckload

SIC Standard Industrial Classification

TIUS Truck Inventory and Use Survey

TOFC/COFC Trailer-on-flatcar/container-on-flatcar

TRB Transportation Research Board

xii

ACKNOWLEDGEMENTS

The authors are grateful to the Center for Logistics Research

at The Pennsylvania State University, which conducted the

shipper survey and ran the Freight Transportation Analyzer (FTA)

computerprogram. The results of the survey and the FTA program

provided the data for the research described in this report. In

particular, the authors would like to thank Mr. Paul Poissant

and Ms. Rosemarie Greaser for compiling and editing the data

from the shipper survey and the output from the FTA; Dr. Gary L.

Gittings, Assistant Professor of Business Logistics, for

providing technical assistance and a description of the survey

methodology; and Dr. Alan J. Stenger, Associate Director of the

Center for Logistics Research, for granting permission to use

his FTA model for this study.

The authors are also grateful to the Federal Highway

Administration, which sponsored the research described in this

report, and in particular, Mr. Jake Jacoby for his valuable

technical guidance.

xiii

EXECUTIVE SUMMARY

While previous studies have indicated that increases in truck

size and weight limits could improve motor carrier productivity,

the question of whether or not freight shippers will also

benefit has not been adequately addressed. It is generally

assumed that competitive conditions in the motor carrier

industry will result in cost savings being passed to shippers in

the form of lower freight rates. Transportation costs, however,

are only one component of shipper total logistics cost.

Warehousing cost, inventory holding cost, order processing

cost, and other categories of business logistics cost may also

change as a result of the less frequent but larger shipments

typically associated with the use of longer combination

vehicles (LCVs). If switching from single trailer truckload

shipments to LCVs causes shipper non-transport logistics costs

to increase more than the savings available from lower freight

rates, then productivity gains may be lost to the firm and the

economy as a whole. This research was undertaken to determine

the net effect of truck size and weight policy changes on

shipper total logistics cost and how these effects might

influence the demand for alternative tractor-trailer

configurations.

One of the more difficult tasks in this study was the

collection of logistics cost data on which to perform the

analysis. Original data of a highly confidential nature was

required to fulfill the study's objectives. Firms are naturally

reluctant to divulge sensitive data which might compromise their

competitive position. Many of the firms contacted in this study

were willing to provide freight flow data for their principal

products; however, even when assured of confidentiality, they

frequently remained either unwilling or unable to specify

critical logistics costs such as order processing cost and

inventory carrying cost. Some of the contacted firms also

lacked the sophisticated logistics management systems necessary

to respond fully to the detailed questions that were asked. As

a result, this research is based on a limited sample of 297

product-specific traffic lane (origin-destination) movements

obtained from a total of 72 companies.

The data on product characteristics, lane volumes,

transportation cost, and other logistics costs gathered in the

shipper survey were entered into a computer program called the

Freight Transportation Analyzer (FTA). Developed by Dr. Alan J.

Stenger of the Pennsylvania State University Department of

Business Logistics, the FTA implements a deterministic economic

order quantity (EOQ) model adapted to incorporate transportation

costs. For each lane observation in the survey dataset, the

FTA calculated the shipper's annual freight, order, and

inventory carrying costs for the shipper's current mode of

transport as well as for two types of LCVs: the Rocky Mountain

double combination consisting of a tractor pulling a 48-ft

(14.6-m) trailer followed by a 28-ft (8.5-m) trailer and the

turnpike double combination which is a tractor pulling two

48-ft (14.6-m) trailers.

A major finding of the study is that, in most cases, use of

LCVs would have a significant favorable impact on the annual

total logistics cost of truckload shippers. Savings in annual

total logistics cost as high as 59 percent for turnpike doubles

and 52 percent for

Rocky Mountain doubles were observed. Average savings depend

on the type of LCV used, gross vehicle weight (GVW) limits, and

the difference between the rates charged by motor carriers for

LCV and single trailer transport. For example, the study

predicted average savings of 32 percent for shippers switching

to turnpike doubles operating at 131,000-lb (59,422-kg) GVW

limits under LCV freight rates that are the same as current

single trailer freight rates. The average savings drop to 21

percent when turnpike double freight charges are 30 percent

higher than current single trailer freight rates.

The greater the flow of a company's product in the lane and

the longer the lane, the more likely it is that a shipper would

benefit from using some type of LCV. When annual lane volumes

are below 5000 cwt (226,795 kg), LCVs tend to increase the

shipper's annual total logistics cost. When annual lane volumes

exceed 25,000 cwt (1,133,975 kg), however, LCVs nearly always

produce significant cost savings.

An excellent indicator of whether or not a truckload shipper

would benefit from switching to LCVs is the ratio of the

shipper's current annual single trailer freight costs to annual

inventory carrying costs. The research indicates that, when

single trailer freight costs are two or more times greater than

the inventory carrying costs, switching from single trailers to

LCVs will in all likelihood greatly reduce the shipper's annual

total logistics cost. On the other hand, when inventory carrying

costs are roughly the same as or greater than the single

trailer freight costs, the chances are good that switching from

single trailers to LCVs will increase the shipper's annual total

logistics cost.

No single variable or combination of variables among the ones

considered in this study appears to be highly effective at

predicting how much or to what degree an individual shipper's

annual total logistics cost would change as a result of

switching to some type of LCV. The influence of product value,

in particular, is much smaller than is commonly expected.

Product value is significant only when annual traffic lane

volumes fall below 15,000 cwt (680,385 kg) or 350,000 ton-mi

(510,650 metric ton-km). Only at low annual shipment volumes do

higher product values significantly increase the chances that

LCV use will increase the shipper's total logistics cost. Other

factors such as annual lane volume and lane distance are good

indicators of whether or not a shipper would benefit from using

LCVs, but they are not highly significant estimators of the

amount that would be saved or lost. Further research with more

detailed shipper data will be needed to produce better logistics

cost models for alternative truck sizes and weights.

Given the significant potential benefits of LCVs to truckload

shippers as indicated by the analysis, the next question is to

what extent would shippers nationwide switch from using single

trailers to some type of LCV. Using data from the 1987 Truck

Inventory and Use Survey (TIUS), an attempt was made to answer

this question. The resulting estimates of nationwide intramodal

diversion to LCVs vary depending on GVW limits, the difference

between LCV and single trailer freight rates, and the minimum

percent logistic cost savings necessary before shippers would

switch to LCVs. For example, given weight limits of 115,000 lb

(52,164 kg) for Rocky Mountain doubles and 131,000 lb (59,422

kg) for turnpike

xvi

doubles and LCV freight rates equal to current single trailer

freight rates, the estimated nationwide intramodal diversion to

LCVs range from 28 percent to 36 percent for weight-limited or

high density freight and from 38 percent to 51 percent for

cube-limited or low density freight. The magnitude of this

diversion is estimated to be 1.1 billion to 1.5 billion truck-

mi (1.8 billion to 2.4 billion truck-km) for high density

freight and 1.S billion to 2.1 billion truck-mi (2.4 billion to

3.4 billion truck-km) for low density freight out of an

estimated base of 4.1 billion truck-mi (6.5 billion truck-km) in

1987. The base figure represents the number of miles operated by

truck tractors pulling loaded single enclosed dry van semi-

trailers with at least two axles on trips at least 200 mi (322

km) long. More research with better data and more

robust logistics cost models is needed to determine whether this

much diversion would actually occur and what the cumulative

nationwide impact on shippers' total logistics cost would be.

Because of the small number of rail boxcar and intermodal

observations in the shipper survey data, it was not possible to

estimate the amount of diversion that might occur from rail to

LCVs. The research indicates, however, that turnpike doubles

operating under higher than existing GVW limits could reduce

shippers' annual total logistics cost enough to induce some

shippers to switch from rail boxcars and intermodal to LCVs.

Additional research is needed to determine how much rail boxcar

and truck-rail intermodal freight might be diverted.

xvii

1. INTRODUCTION

BACKGROUND

Allowable limits on truck size and weight has been a recurring

issue at both the State and Federal level since the earliest

days of motor carriers and public roadbuilding. Lately the issue

has been raised once again. The Surface Transportation

Assistance Act of 1982 increased the allowable width and length

of tractor-trailer combinations and permitted the use of double

trailer combinations on Interstates and designated Federal aid

primary highways. The Surface Transportation and Uniform

Relocation Assistance Act of 1987 provided for studies to be

undertaken that would investigate the effects of even longer and

heavier tractor-trailer configurations. The American Association

of State Highway and Transportation Officials (AASHTO) also

requested a study of new approaches to regulating truck size and

eight. The resulting studies examined the effects of longer

combination vehicles (LCVs) on motor carrier productivity,

highway safety, highway capital and maintenance costs, and the

transportation costs of shippers and receivers.(1,2,3) Although

some of these studies have considered the effects of LCVs on

shippers' transportation costs, most have not explicitly looked

at each component of a shipper's total logistics cost and how it

might affect the demand for l CV service.

Recent preliminary research provided some initial insight

into the relationship between a shipper's logistics costs and

the economic advantages of LCVs.(4) The study revealed that the

net benefits of larger truck sizes and weights to shippers were

sensitive to product value, length of haul, and lane volume. In

particular, the study found that the user's inventory carrying

costs comprised an increasingly larger proportion of the total

logistics cost as product value increased and as shipping

distance and lane volume decreased. The results indicated that

certain combinations of product value, shipment distance, and

annual lane volume can increase inventory carrying costs to the

point where they offset the transportation cost savings offered

by longer combination vehicles.

The results of the above study were based on data from a

small sample of shippers and covered only a few types of

products. A need existed to extend this firm-level research to a

broader range of industries, product values and densities,

shipment distances, and annual volumes. There also existed a

need to develop, test, and implement models to scale the

research findings from the level of the individual shipper or

firm to the national level so as to provide nationwide estimates

of the effects of Federal truck size and weight policy options.

The research described in this report was undertaken to satisfy

these needs.

RESEARCH OBJECTIVES

The main objectives of this research were the following:

. Analyze the effects of Federal truck size and weight

policy options on shippers'component logistics costs.

. Estimate the nationwide diversion of freight from

conventional, single trailer truck configurations and

from other modes to LCVs, based on the results of the

above analysis.

The study attempted to complement the knowledge gained from

previous studies by looking explicitly at how different truck

sizes and weights can affect the component logistics costs of

individual shippers. The intent was to acquire a broader

perspective on the productivity gains and economic

implications of changes in Federal truck size and weight

policy.

ORGANIZATION OF THE REPORT

The research findings were based on a mail and telephone

survey of shippers and a mathematical model for computing a

firm's logistics costs. Chapter 2 discusses what information

was collected from shippers and how the sample was obtained.

It also describes the Freight Transportation Analyzer (FTA), a

computer program which implements the model for calculating

logistics costs. Chapter 2 also contains statistics on the

size and characteristics of the sample.

Chapter 3 presents the results of the FTA model for

those companies in the survey sample who shipped their

products in single semi-trailers of various sizes. It

describes the overall effects of LCVs on the total logistics

cost of these shippers as well as the influence of such

factors as product value, shipment distance, annual lane

volume, and combinations of these factors. It also provides

estimates of LCV diversion rates under various situations.

Chapter 4 covers the results of the FTA model for

companies in the survey sample that shipped via rail boxcar or

trailer-on-flatcar (TOFC) intermodal service. Because of the

much smaller sample sizes involved, the discussion is not as

extensive as that of Chapter 3. Nevertheless, Chapter 4

discusses the overall effects of LCVs on the logistics costs

of rail and intermodal shippers as well as the influence of

rail and intermodal freight rates on intermodal diversion to

LCVs.

In Chapter 5, the results of the firm-level analysis are

applied to data from the 1987 Truck Inventory and Use Survey

(TIUS) to estimate the nationwide diversion of truck traffic

to LCVs. Chapter 5 describes the estimation methodology and

compares the resulting estimates with those from a previous

study.

Chapter 6 summarizes the major findings and limitations of

the research.

Throughout this report the terms "rate", "freight rate", and

"freight charge" are used interchangeably. They refer to a

carrier's price for shipping a certain quantity of cargo

between an origin and a destination. In practice this price

could be quoted per unit, per hundredweight (cwt), or per mile.

In this report, unless noted otherwise, carrier rates or charges

are specified in dollars per mile for a truckload or a

boxcarload shipment. The terms "transportation cost" and

"freight cost" refer to the total amount that a shipper pays for

a shipment between an origin and a destination. It is equal to

the carrier's rate per mile times the shipment distance or the

carrier's rate per hundredweight times the weight of the

shipment. A shipper's annual freight cost is equal to the annual

number of shipments times the carrier's freight charge per

shipment.

2. DATA COLLECTION AND LOGISTICS COST ANALYSIS METHODOLOGY

INTRODUCTION

The firm-level analysis was based on data from a mail

and telephone survey of companies in selected industries. The

information provided by the surveyed companies was then

entered into a computer program called the Freight

Transportation Analyzer (FTA) to calculate component logistics

costs for alternative types of longer combination vehicles as

well as for the shipper's current primary mode of

transportation.

This chapter addresses the survey methodology and the

FTA. It identifies the kinds of information requested from the

surveyed companies, describes how the sample of shippers was

obtained, and presents statistics on the size and

characteristics of the resulting sample. The discussion then

turns to the FTA, including a description of its input and

output. The chapter concludes with a description of the size,

structure, and characteristics of the resulting FTA datasets.

SHIPPER SURVEY

Information Collected

The survey was designed to collect product information,

freight flow data, and logistics cost information from each

company.

Shippers were first asked to select and describe two of

their company's primary products or product groups. The

following data were requested for each principal product:

. Standard Industrial Classification (SIC) code or a

written description of the product.

. Unit weight of the product in pounds where a unit

represented the wholesale or shipping package such as a

case, sack, bale, or tote.

. Cost of the product per pound or unit.

. Whether the product would cube-out or weigh-out; that

is, whether the product would use up the cubic capacity

of a 48-ft (14.6-m) long, 102-in (2.6-m) wide semi-

trailer before reaching the maximum weight limit or

whether it would reach the weight limit before filling

up the cargo space of such a trailer.

The surveyed companies were then asked to select a

representative short-distance (less than 500 mi [805 km]),

medium-distance (500 to 1,000 mi [805 to 1,609 km]), and long-

distance (over 1,000 mi [1,609 km]) lane or origin-destination

pair for each of the two principal products and to provide the

following freight flow data for each combination of lane and

product:

. Origin city.

. Destination city.

. Lane distance in miles.

. Annual volume in pounds or number of shipments.

. Primary mode or type of carrier.

. Freight rate.

. Typical size of trailer, container, or rail boxcar used.

. Average weight per shipment.

In selecting representative lanes and freight flows, shippers

were asked to observe the following guidelines:

. Consider only outbound movements where the shipper

controlled the volume and route. such as shipments between

a plant and a distribution center.

. Include only direct movements, not movements involving

stop-offs or special handling.

. Include only movements in trailers, intermodal containers,

or rail boxcars.

. Exclude shipments involving specialized equipment such

as bulk hoppers or tankers.

. Include only full-sized shipments, defined as having one

or more of the following characteristics:

. At least 14,000 pounds (6,350 kg).

. At least 18 pallets or slipsheets.

. At least 50 percent of the cubic capacity of a 40-ft

(12.2-m) long trailer or container or a rail boxcar.

Finally, shippers were asked to specify their company's

average cost of processing an order or shipment and their

company's inventory carrying cost. The latter was reported as

a percentage of inventory value. Order costs involve the cost

of handling the paperwork or electronic data interchange for

an order or a shipment. Inventory carrying costs typically

cover the cost of capital, warehousing, taxes, insurance,

depreciation, and obsolescence. These two cost categories,

along with transportation costs, form the primary components

of shippers’total logistics cost.

Data Collection Methodology

The nature of the data to be collected posed some

challenging problems in devising an effective data collection

method. First, cost information is closely guarded in most

firms. Some companies go so far as to impose corporate

policies against disseminating any sensitive cost information.

Second, four types of cost information were required: product

costs, inventory carrying costs, order costs, and freight

transportation costs (as well as traffic volume data).

Information on all four of these cost categories is not

necessarily maintained or controlled by one individual or even

one functional unit within an organization. To provide all of

the requested information, a company contact would in many

cases have to coordinate with other functional groups, greatly

increasing the likelihood of either an incomplete response or

no response at all.

While a broad mix of products, company sizes, and

shipment volumes was desirable, including all possible

products and shipping patterns in the sample would have

greatly increased the time and cost of running the FTA program

and analyzing the results. Therefore, an effort was made to

target the survey at companies matching the following profile:

. Ship in truckload or carload quantities.

. Use dry transportation equipment such as dry vans,

containers, or boxcars instead of flatbeds, tankers,

reefers, or bulk hoppers.

. Make direct shipments with little or no special handling

requirements.

Because of these and other obstacles, three separate

data collection efforts were eventually required to obtain an

adequate sample size. Each effort employed a different

sampling frame.

Phase One

The initial effort involved a mail and telephone survey,

using as the sampling frame the 1992 edition of The Official

Directory of Industrial & Commercial Traffic Executives

(commonly referred to as the Bluebook). Companies were randomly

selected. To increase efficiency and ensure broad geographic

coverage, samples were drawn from five base business regions:

. East: eastern New York, northern New Jersey, and

southern Connecticut.

. South: Georgia and southern South Carolina.

. Midwest: northeastern Illinois, southern Wisconsin,

northwestern Michigan, and northern Indiana.

. Southwest: Texas.

. West: southern California.

The sample was further stratified by two-digit SIC code to

ensure a broad mix of basic commodities.

The randomly chosen companies were contacted by

telephone and questioned about their current shipment patterns

and their willingness to participate in the research.

Approximately 3,700 phone calls were made in the effort to

screen potential participants. Eligible companies received a

five-page questionnaire along with a cover letter and a

postage-paid, self-addressed return envelope. Follow-up phone

calls were made when necessary two to three weeks after the

initial contact to answer questions and improve the response

rate.

The results of the initial data collection effort were

as follows:

. 731 randomly selected companies were contacted

and screened for eligibility.

. 251 companies were sent a questionnaire.

. 11 companies completed the survey satisfactorily,

resulting in a response rate of only 4.38 percent.

Several factors contributed to the extremely low

response rate. A major factor was the complexity of the

information being requested, especially the logistics cost

information. Many companies were able to provide freight flow

data but were either unwilling or unable to specify their

order processing and inventory carrying costs. Many of the

small firms lacked the sophisticated logistics management

necessary to interpret and respond properly to the detailed

questions about logistics. Use of a general business directory

for the sampling frame was another factor. As much as 20 to 30

percent of the information in such directories may be

outdated. In addition, rules for inclusion in general business

directories can be very broad, whereas the rules for

eligibility for the survey were quite restrictive.

Consequently, many of the randomly selected firms were

ineligible for the survey because they generally shipped in

less-than-truckload (LTL) quantities. A third factor was the

Commodity Flow Survey (CFS) being conducted at the same time

by the U.S. Bureau of the Census. Many of the contacted firms

were heavily involved in providing shipment data for the CFS

and, consequently, were unable to take the time required to

respond to this survey.

Phase Two

Because the survey was conducted by the Center for

Logistics Research at The Pennsylvania State University, a

directory of alumni with degrees in business logistics was

readily available. This sampling frame was therefore used in

the second data collection effort. Concentrating on business

logistics alumni offered the following potential advantages:

. Penn State's large business logistics network allowed

access to a wide number and variety of firms.

. Contacting individuals with advanced logistics training should greatly improve comprehension of the survey questionnaire.

. Loyalty to one's alma mater should increase the

willingness to respond.

. The information in the alumni directory was of

higher quality than that in many standard lists and

directories.

As in the first data collection effort, alumni at

prospective firms were contacted by telephone and questioned

about their company's current shipment patterns and

willingness to participate in the research. Eligible companies

then received the survey questionnaire along with an

introductory letter. Follow-up phone calls were made two to

three weeks after the initial contact to answer questions and

solicit cooperation.

The results of the second data collection activity were

as follows:

. 271 companies were contacted and screened for

eligibility.

. 187 companies were sent a questionnaire.

. 52 companies completed the survey satisfactorily,

resulting in a response rate of 27.8 percent.

Phase Three

The third data collection effort involved mailing the

survey questionnaire to approximately 500 companies chosen

from the 1993 edition of the Council of Logistics Management

Membership Roster. In this case no effort was made to screen

the companies for eligibility except to ensure that no

companies contacted during either of the previous data

collection efforts were included in the mailing. Follow-up

calls, however, were made two to three weeks after the initial

mailing.

Only nine of the 500 companies responded satisfactorily,

a response rate of only 1.8 percent.

Sample Size and Characteristics

The final shipper survey sample included 72 companies,

and the resulting dataset contained 297 traffic lane

observations. Table 1 shows a breakdown of these observations

by type of product. Shipments of food and kindred products

(SIC 20) and shipments of chemicals and allied products (SIC

28) together accounted for nearly half (47.5 percent) of the

total number of lane observations. Paper and allied products

(SIC 26) along with stone, clay, and glass products (SIC 32)

contributed another 18 percent of the observations. Product

values ranged from $0.0045 per lb ($0.01 per kg) to $75.00 per

lb ($165.35 per kg) with an average of $4.14 per lb ($9.13 per

kg). Table 2 shows the distribution of transport modes and

types of equipment among the observations. In over 80 percent

of the cases, trucks were the principal mode used in the

traffic lane. About one out of every six observations involved

rail transport, either trailer-on-flatcar (TOFC) or boxcar.

Inventory carrying costs ranged from 2.0 percent of

inventory value to 140 percent. The average inventory carrying

cost was 18.1 percent of inventory value. Of the 72 companies

responding to the survey, 13 were unable to provide any

information on their inventory carrying costs. Some companies

stated that their current product costs and interest rates

were so low that inventory carrying costs were not considered

in making daily transportation decisions.

Fixed order costs varied between $2.20 and $320.00 per

shipment. The average was $44.53. Only 53 of the 72 companies

in the sample were able to specify their fixed order costs.

Click HERE for graphic.

FREIGHT TRANSPORTATION ANALYZER

The Freight Transportation Analyzer (FTA) is a personal

computer-based program that calculates not only the freight

costs of various transportation options but the total logistics

cost as well, including the cost of carrying inventories at the

origin and destination. It essentially implements a

deterministic economic order quantity (EOQ) model adapted to

incorporate transportation costs.

The FTA program takes as input data about the product, the

transportation alternatives, related costs at the origin and

destination, and the inventory management system at the

destination. Product data include unit weight, annual demand for

the product at the destination, and product value at the origin.

Transportation alternatives are described in terms of maximum

quantity per shipment, fixed charges per shipment, and average

transit time. Cost data include fixed order costs per shipment

andinventory carrying costs in transit and at the destination.

Input data on the inventory management system at the destination

include the administrative lead time, the inventory review

period, and the frequency at which demand forecasts are updated.

In addition to these data, the FTA also accepts input for many

other variables including the accuracy of a firm's demand

forecasts, transit time variability, in-transit damage factors,

and production scheduling. These latter variables, however,

were not considered important to the objectives of this study.

The output of the FTA program is a detailed breakdown of a

shipper's total annual logistics cost associated with each

transportation alternative for a given product and traffic lane.

In this study the detailed costs were summarized into three

components:

. Annual freight cost.

. Annual fixed order cost.

. Annual inventory carrying cost.

The FTA was run on each product-lane observation for which

complete information was provided. Because some companies could

not provide data on their inventory carrying cost or their fixed

order cost, not all 297 observations could be used. For each

complete product-lane observation, the FTA calculated the

freight, order, and inventory carrying costs for the shipper's

current mode of transport as well as for each of the following

two types of longer combination vehicle:

. Rocky Mountain double, which is a tractor pulling a 48-ft

(14.6-m) trailer followed by a 28-ft (8.5-m) trailer.

. Turnpike double, which is a tractor pulling two 48-ft

(14.6-m) trailers.

For product-lane observations involving shipments that

typically weigh-out, two FTA cases were generated, one for

each of the two gross vehicle weight (GVW) scenarios defined

in table 3. This table shows the GVW limit and typical payload

for each type of LCV in each scenario. The difference between

the GVW limit and typical payload is greater than the tare

weight of the LCV because it is generally not possible to

completely fill the cubic capacity of both trailers. Under the

low weight capacity scenario, the current GVW limit of 80,000

lb (36,288 kg) is retained for both types of LCV. Because of

the extra tare weight of a 48-ft (14.6-m) trailer over a 28-ft

(8.5-m) trailer, the typical payload for a turnpike double

under this scenario is less than the typical payload for a

Rocky Mountain double. The GVW limits are higher under the

high weight capacity scenario and are set to take advantage of

the additional cubic capacity offered by each type of LCV. The

higher GVW limits, however, are not in proportion to the

additional cubic capacity. They are affected by the Federal

Bridge Formula, which is designed to protect bridges,

especially those with relatively long spans, from the

additional weight. In both GVW limit scenarios, the base case

was the shipper's current mode of transport under existing GVW

limits.

Table 3. Weight capacity of longer combination vehicles under

two gross vehicle weight (GVW) limit scenarios.

Click HERE for graphic.

metric conversion: l lb = 0.4536 kg

For lane observations involving shipments that typically

cube-out, the number of FTA runs made for each observation

depended on the current payload weight. Only one FTA run was

necessary if the current payload was less than 38,000 lb

(17,237 kg). Because cubic capacity rather than GVW is the

limiting factor for these shipments, the results of the FTA

program would have been the same for each GVW scenario.

Therefore, only one FTA run was necessary for each of these

lane observations, and its output was used for both GVW

scenarios. The cubic capacity ratios shown in table 4 were

used to determine shipment size and, therefore, the number of

shipments required to accommodate the annual lane volume under

each alternative mode of truck transport. Most of the

observations of shipments that typically cube-out had current

payloads of less than 38,000 lb (17,237 kg). For the few

observations of shipments that typically cube-out and weigh

38,000 lb (17,237 kg) or more, two FTA runs were made. One run

used the GVW limits and typical payloads for the low weight

capacity scenario. The other run was based on either the GVW

limits and typical

payloads for the high weight capacity scenario or the cubic

capacity ratios shown in table 4, depending on whether GVW or

cubic capacity was the limiting factor.

Table 4. Cubic capacity of longer combination vehicles

compared to single trailers of various sizes.

Click HERE for graphic.

Metric conversion: 1 ft3 = 0.028 m3

The payload and cubic capacity guidelines shown in

tables 3 and 4 were compiled from numerous sources, including

the Western Highway Institute, the American Trucking

Associations, interviews with shippers and carriers, and

previous studies of longer combination vehicles.

A number of companies in the shipper survey reported a

range of product values rather than a single value. For these

product-lane observations, one or two FTA runs were made based

on the high product value and another one or two runs were

made using the low product value. The number of FTA cases

generated for each product value depended on the weight of the

payload and whether the shipment typically cubed-out or

weighed-out, as described above.

If a company reported using more than one mode or

trailer size in a lane, the one chosen for the FTA run was the

mode or trailer size involved in transporting the bulk of the

annual lane volume.

An important input to the FTA is the carrier's rate or

freight charge for each alternative mode. Shippers provided

this information in the survey for their current mode. For

each of the LCV alternatives, two LCV rate assumptions were

considered for shipments currently made by truck. The first

assumption was that shippers would pay the same rate for each

type of LCV that they were currently paying for single trailer

transport. This

assumption is not unrealistic. In a recent survey of motor

carriers who operate LCVs, SO percent of the respondents

stated that LCV operation had lowered their typical rates, and

46 percent said there was no significant difference.(5) The

FTA was therefore run on the basis of this first assumption.

The second assumption was that LCV rates would be higher,

reflecting the higher costs per mile of operating LCVs. This

assumption was implemented by a derived set of factors which

were applied to the annual freight costs calculated by the FTA

under the first LCV rate assumption. Table 5 indicates these

factors and how they were derived. The operating costs per

loaded mile were taken directly from a previous study.(6)

These costs were based on the assumption that 15 percent of a

vehicle's mileage is empty haul. The gross vehicle weights

shown in the table were the ones used in the previous study.

Although they differ slightly from the gross vehicle weights

assumed for the high and low weight capacity scenarios, the

costs that vary with weight do not change rapidly enough to

alter the derived set of factors significantly.

Table 5. LCV rate adjustment factors based on a comparison of

operating costs per loaded mile.

Click HERE for graphic.

metric conversion: 1 ft = 0.3 )48 m; 1 mi = 1.609 km; 1 lb =

0.4536 kg

The second LCV rate assumption affected only the

computation of the total annual freight cost for each FTA run.

Calculation of annual order and inventory carrying costs by

the FTA does not depend on the LCV rate. Consequently, there

was no need to rerun the FTA for the second LCV rate

revised annual freight costs were then added to the order and

inventory carrying costs from the FTA runs to produce total

logistics costs under the second LCV rate assumption.

The LCV rate assumptions described above applied only to

product-lane observations where the current mode was truck.

For lanes in which the product was currently shipped by rail

boxcar or rail-truck intermodal, an LCV rate of $1.30 per mi

($0.81 per km) was used to run the FTA. Under the assumption

that LCV operators will charge the same rate as currently

charged for single trailers, this rate turned out to be perhaps

a little too high. Nearly all of the rail and intermodal

observations involved either medium lane or long lane distances.

The average single trailer freight charges paid by shippers in

the survey were $1.24 per mi ($0.77 per km) for medium lane

shipments and $1.14 per mi ($0.71per km) for long lane

shipments. On the other hand, under the assumption that LCV

operators will charge higher rates for LCV service because of

the higher operating costs involved, the assumed LCV rate of

$1.30 per mi ($0.81 per km) was perhaps too low for some

situations. After applying the rate adjustment factors in

table 5 to the current single trailer rates charged shippers

in the survey, the results showed that the average rate for

Rocky Mountain doubles for medium lane distances would be

$1.36 per mi ($0.85 per km) under the low weight capacity

scenario and $1.42 per mi ($0.88 per km) under the high weight

capacity scenario. For long lane shipments the respective

rates would be $1.25 per mi ($0.78 per km) and $1.31 per mi

($0.81 per km). The average rates for turnpike doubles were

determined to be $1.48 per mi ($0.92 per km) and $1.61 per mi

($1.00 per km) for medium lane shipments under the low and

high weight capacity scenarios, respectively, and $1.36 per mi

($0.85 per km) and $1.48 per mi ($0.92 per km) for long lane

shipments under the two GVW scenarios. The originally assumed

LCV rate of $1.30 per mi ($0.81 per km), therefore, appeared

to be a reasonable compromise. Because of the small number of

rail and intermodal shippers in the sample, no effort was made

to rerun the FTA for different LCV rates for these cases.

Table 6. Breakdown of FTA observations by current principal

mode.

Click HERE for graphic.

Altogether some 347 runs of the FTA were made. The

output was organized into two datasets, one for each GVW

scenario. Each dataset contained 228 cases or observations.

Table 6 shows the breakdown of these cases by current

principal mode. Each observation consisted of the annual

freight cost, order cost, inventory carrying cost, and total

logistics cost for the current mode, Rocky Mountain double

LCV, and turnpike double LCV as determined by the FTA.

3. EFFECTS OF LCV USAGE ON THE LOGISTICS COSTS

OF TRUCK SHIPPERS

INTRODUCTION

This chapter presents the results of the FTA runs for

shipments currently handled by tractor and single trailer. It

begins by addressing the overall effect of LCV use on the

total logistics cost of current motor carrier users, showing

how the cost consequences of LCV use can vary by type of

LCV,GVW limits, and LCV rates relative to single trailer

rates. It then looks at several individual factors and

combinations of factors that could influence how much of an

effect LCV use would have on a truck shipper's total logistics

cost. Based on these findings, the chapter then considers how

many shippers might switch from single trailer truck hauls to

some type of LCV.

OVERALL EFFECTS

The FTA results indicated that, in a large majority of

cases, companies would reduce their total logistics cost by

switching from single trailer truck transport to some type of

LCV. The extent to which a company's total logistics cost for

a given product in a given lane would change as a result of

switching depended on the type of LCV,GVW limits, and LCV

rates relative to single trailer rates. Table 7 summarizes the

FTA results with respect to these three parameters.

Two of the more noteworthy findings in table 7 are the

large average and median percent reductions in total logistics

cost associated with LCV usage and the wide range in the

percent change in costs that can result from switching to

LCVs. For example, under the high weight capacity scenario

with LCV rates the same as current single trailer truck rates,

the FTA calculated that use of Rocky Mountain doubles would

lead to cost savings of at least 33 percent in half of the

product-lane observations. Under the same situation, use of

turnpike doubles would lead to cost reductions of at least 42

percent in half of the cases. For the vast majority of

product-lane observations, switching to some type of LCV would

reduce total annual logistics cost. The average reduction in

total logistics cost for Rocky Mountain doubles ranged from 17

percent to 27 percent, depending on GVW limits and LCV rates

relative to single trailer truck rates. For turnpike doubles

the range in average cost reduction was between 13 percent and

32 percent. In some cases cost savings as high as 45 to 52

percent were predicted by the FTA for Rocky Mountain doubles

and 48 to 59 percent for turnpike doubles. At the other

extreme, significant cost increases were also possible. The

FTA results revealed cases where switching to Rocky Mountain

doubles would increase the shipper's total annual logistics

cost by as much as 42 percent and switching to turnpike

doubles would increase it by as much as 73 percent.

Table 7. Overall effect of LCVs on the total logistics cost of

single trailer truck shippers by type of LCV,GVW limits, and

relative LCV rates.

Click HERE for graphic.

Notes: Negative percentages imply an increase in total

logistics cost. Percentages in the rightmost column are

based on a sample size of 176 cases.

Higher GVW limits tended to increase the economic benefits of

LCVs. The average and the median percent cost reduction

resulting from the use of LCVs was generally higher under the

high weight capacity scenario than under the low weight capacity

scenario. The importance of higher GVW limits was especially

noticeable for turnpike doubles. In going from an 80,000-lb

(36,288-kg) to a 131,000-lb (59,422-kg) GVW limit, the average

percent cost reduction for turnpike doubles increased between 7

and 9 percentage points, depending on the relative LCV freight

rate. The increase in the median percent cost reduction was

even larger at around 20 percentage points. For Rocky Mountain

doubles, on the other hand, the effect of going from an

80,000-lb (36,288-kg) GVW limit to a 115,000-lb (52,164-kg)

limit was an increase in the average percent cost reduction of

only about 2~/: to 4 percentage points.

Higher LCV rates had a noticeable but not tremendously large

dampening effect on the generally favorable economic

consequences of LCV use. Compared to the situation in which the

LCV rate in each case was the same as the shipper's current

single trailer truck rate, higher LCV rates based on higher

operating costs per mile tended to reduce the average

percent cost reduction by around 5 to 12 percentage points,

depending on LCV type and GVW limits. Again the effect was

stronger for turnpike doubles than for Rocky Mountain doubles.

Higher LCV rates also increased the number of cases where a

shipper would experience an increase in total annual logistics

cost by switching to LCVs, but only to a fairly small extent.

The relative attractiveness of the alternative truck

configurations greatly depended on GVW limits. Table 8 shows the

distribution of lane observations by type of truck configuration

resulting in the lowest total logistics cost. When GVW limits

were kept at their current level of 80,000 lb (36,288 kg), Rocky

Mountain doubles held a slight edge over turnpike doubles. At

current GVW limits, turnpike doubles are virtually worthless for

truckload shipments that typically weigh-out in single trailers.

The FTA results indicated, however, that for low-density

products that normally cube-out in single trailer truckload

shipments, turnpike doubles could be advantageous even under

current weight limits. Turnpike doubles were by far the most

cost-effective configuration under the high weight capacity

scenario where shipments that normally weigh-out in single

trailers can take advantage of the additional cubic capacity

offered by twin 48-ft (14.6-m) trailers.

Table 8. Truck configuration resulting in the lowest total

logistics cost by GVW limits and relative LCV rates.

Click HERE for graphic.

Higher LCV rates relative to current single trailer rates also

influenced which configuration was the most economical in terms

of total logistics cost. Compared to GVW limits, however, the

effect was relatively modest. In general it was to increase the

attractiveness of single trailers and Rocky Mountain doubles at

the expense of turnpike doubles' particularly under the high

weight capacity scenario.

INFLUENCE OF VARIOUS COST, PRODUCT, AND LANE VARIABLES

The FTA results presented in the previous section showed

that use of LCVs would lower the total logistics cost of the

surveyed shippers in the vast majority of cases. The results,

however, also showed tremendous variation in the degree to

which the total logistics cost would change. In an effort to

explain this wide variation, a number of variables related to

costs and to product and traffic lane characteristics were

examined individually and in combination with each other to

determine what effect they had on the percent change in total

logistics cost.

Ratio of Freight to Inventory Costs

Switching from single trailer truck transport to an LCV

entails trading off higher inventory carrying costs against

lower shipping costs. The degree to which using LCVs raises or

reduces a firm's total logistics cost could depend on the

comparative magnitude of the firm's freight and inventory

carrying costs using single trailers. If the inventory

carrying cost of a commodity is much smaller than the cost of

shipping it in single trailer truckload quantities in a given

traffic lane, then switching to an LCV will probably not

increase the inventory carrying cost more than it decreases

the freight cost. The net result would be a reduction in total

logistics cost. On the other hand, if the inventory carrying

cost is already about the same as or greater than the cost of

shipping, then switching from single trailers to LCVs would

likely increase the inventory carrying cost more than it

reduces the freight cost, resulting in an increase in the

total logistics cost.

Figures 1 and 2 support the above hypotheses. Each

figure is a plot of the percent reduction in total logistics

cost as a result of switching to LCVs versus the ratio of

freight cost to inventory carrying cost for single trailer

transport. Figure 1 shows the relationship under the low

weight capacity scenario while figure 2 corresponds to the

high weight capacity scenario. Both figures are based on the

assumption that LCV rates will be higher than single trailer

rates; however, the patterns displayed in these plots were the

same under the assumption that LCV rates would not differ from

current single trailer truckload rates.

In the FTA results, the percent change in total

logistics cost was highly sensitive to changes in the

freight-to-inventory cost ratio for ratios below approximately

4.0. LCV usage always resulted in a lower total logistics cost

when the freight cost for single trailer shipments was at

least twice as large as the inventory carrying cost. The ratio

at which LCVs went from reducing to increasing the total

logistics cost was generally between 1.5 and 1.7. In most

cases where the ratio was less than 1.0, meaning that the

inventory carrying cost was already greater than the single

trailer freight cost, use of LCVs resulted in a higher total

logistics cost. The more the ratio dropped below 1.0 (that is,

the more the inventory carrying cost increased over the

freight cost), the greater the negative impact of LCV usage on

the total logistics cost.

Click HERE for graphic.

As figures 1 and 2 indicate, freight-to-inventory cost

ratios in the FTA datasets were generally quite high. The

median ratio was 10.28, and in 81.25 percent of the cases, the

current single trailer freight cost was at least twice as

large as the inventory carrying cost. There were only 12 cases

(6.8 percent of the total) in which the inventory carrying

cost exceeded the single trailer freight cost.

Figures 1 and 2 suggest an obvious way by which a firm

could determine whether or not and to what extent it might

benefit from using LCVs. The procedure is to determine the

firm's inventory carrying cost and compare it to the cost of

shipping via single trailer truckloads in the traffic lane. If

the shipping cost is greater than the inventory carrying cost

by a factor of two or more, then use of LCVs would most likely

be justified from the standpoint of lowering the firm's total

logistics cost. Of course, market-related factors as well as

others besides logistics costs may influence whether or not a

firm will consider using longer combination vehicles.

Product Value

Given that LCVs are likely to be inappropriate in

situations where the inventory carrying cost is nearly the

same as or larger than the transport cost, it would seem that

anything which increases the cost of storing a particular

product would tend to reduce the attractiveness of LCVs. Many

factors determine a firm's cost of holding inventory. They

include interest rates, insurance, property taxes, capital,

warehousing, depreciation, and obsolescence. Many of these

factors, in turn, are affected by product value. For example,

high-valued products often require special handling and

storage in climate controlled warehouses. Thus, product value

can greatly influence inventory carrying cost, although the

relationship may sometimes be subtle and indirect.

Nevertheless, the FTA results were used to test the hypothesis

that, as product value increases, the positive effect of LCVs

on total logistics cost decreases.

The product-lane observations in the FTA datasets

encompassed a wide range of product values, varying from a

minimum of $0.02 per lb ($0.04 per kg) to a maximum of $75.00

per lb ($165.35 per kg). Roughly one-third of the observations

had product values under $1.00 per lb ($2.20 per kg) while

another third had product values at or above $3.00 per lb

($6.61 per kg). Thus, the sample contained an adequate mix of

high- and low-valued commodities with which to test the above

hypothesis.

Table 9 shows the observed relationship between product

value and percent cost reduction associated with use of LCVs

for different LCV types, GVW limits, and relative LCV rates.

Except for turnpike doubles under the low weight capacity

scenario, the signs of the correlation coefficients generally

supported the hypothesis that higher product values are

associated with less favorable results from using LCVs.

However, the magnitudes of the coefficients were small,

indicating that the relationship is a fairly weak one. Product

value

by itself, therefore, did not help to explain much of the wide

variation in the impact of LCV use on total logistics cost.

Table 9. Correlation between product value and the percent

reduction in total logistics cost resulting from

LCV usage.

Click HERE for graphic.

Inventory carrying cost, expressed as a percentage of

inventory value, varied between 2 percent and 140 percent in

the FTA datasets. This variable was multiplied by product

value to form a composite variable which was then correlated

with the percent reduction in total logistics cost from using

LCVs. The results are shown in table 10. The magnitudes of the

correlation coefficients were once again very small,

indicating very little correlation between the composite

variable and the cost effects of LCV usage. Moreover, the

signs of the coefficients in this case were inconsistent. They

indicated a negative correlation for the high weight capacity

scenario and a positive correlation for the low weight

capacity scenario. As was the case for product value alone,

the result of multiplying product value by the inventory

carrying cost expressed as a percentage of inventory value was

not very helpful in explaining the FTA results.

Annual Lane Volume

As the demand for a product rises in a given traffic

lane and the volume shipped to meet the demand increases,

freight costs tend to account for a larger proportion of the

total logistics cost. Conversely, as lane volume decreases,

inventory costs tend to increase as a percentage of total

logistics cost.(4) Heavy annual lane volumes, therefore,

should favor the use of LCVs.

Table 10. Correlation between percent cost reduction from

using LCVs and the composite variable formed by

multiplying product value by the inventory carrying

cost expressed as a percentage of inventory value.

Click HERE for graphic.

Annual lane volumes for truck shipments in the FTA

datasets varied considerably. The smallest volume was 351 cwt

(15,921 kg), while the largest was 2,970,000 cwt (134,716,230

kg). Half of the observations involved annual lane volumes of

25,000 cwt (1,133,975 kg) or more.

The FTA results indicated that annual lane volume does

influence the effect of LCVs on total logistics cost. Table 11

shows the percent reduction in total logistics cost from using

LCVs at three levels of annual lane volume. Below annual

volumes of 5,000 cwt (226,795 kg), the effectiveness of LCVs

dropped sharply. In fact, the majority of cases in which LCV

usage increased the total logistics cost involved annual lane

volumes of less than 5,000 cwt (226,795 kg). The average

effect of LCVs at the lowest volume level ranged from a 3.39

percent increase in total logistics cost for turnpike doubles

operating under high GVW limits and higher rates relative to

single trailer rates to a 9.62 percent decrease in total

logistics cost for Rocky Mountain doubles operating under low

GVW limits and rates equivalent to existing single trailer

rates. At volumes above 5,000 cwt (226,795 kg), on the other

hand, LCVs decreased total logistics cost by an average of 16

percent to 42 percent, depending on LCV type, GVW limits, and

relative LCV rates. Under the low weight capacity scenario,

the average impact of LCVs did not change much at volumes in

the middle and upper ranges. Under the high weight capacity

scenario, however, the average percent reduction in total

logistics cost was about 7 percentage points higher for Rocky

Mountain doubles and 9 to 10 percentage points higher for

turnpike doubles at volumes over 25,000 cwt (1,133,975 kg)

compared to volumes between 5,000 and 25,000 cwt (226,795 to

1,133,975 kg).

Table 11. Average percent reduction in total logistics cost

from use of LCVs for different levels of annual

lane volume.

Click HERE for graphic.

Metric conversion: 1 cwt = 45.359 kg

While the FTA results showed a logical overall

relationship between annual lane volume and the

cost-effectiveness of LCVs, the relationship was not a simple

linear one. Moreover, the ability of annual lane volume to

predict the actual amount of change in total logistics cost

resulting from the use of LCVs was extremely weak. Table 12

shows the correlation between annual lane volume and percent

cost reduction. In general, the correlation coefficients were

very low. These coefficients along with the results shown in

table 11 suggest that annual lane volume is a good indicator

of the direction in which total logistics cost will change as

a result of using LCVs but is a poor indicator of how much of

a change will occur.

The correlation coefficients in table 12 were much

higher for the high weight capacity scenario than for the low

weight capacity scenario. In addition, the coefficients for

turnpike doubles operating under low GVW limits indicated a

negative correlation between annual lane volume and percent

cost reduction. These results further illustrate the relative

inefficiencies of operating LCVs in general and turnpike

doubles in particular under current GVW restrictions. Much of

the additional cargo space cannot be utilized. As a result, as

the volume in the traffic lane increases, more LCV shipments

are required than would be necessary under higher GVW limits.

Consequently, freight costs also rise, in some cases more

rapidly than they would if single trailers were employed.

Table 12. Correlation between annual lane volume and the

percent reduction in total logistics cost

resulting from LCV usage.

Click HERE for graphic.

Lane Distance

Generally, the farther a product is shipped, the lower

the freight cost is per mile or per hundredweight. However,

the total freight cost tends to contribute to a greater share

of the total logistics cost as lane distance increases.

Conversely, as shipping distances decrease, inventory carrying

costs tend to account for a higher proportion of the total

logistics cost.(4) LCVs, therefore, should become more

cost-effective as shipping distances increase.

Table 13 shows the average percent reduction in total

logistics cost from LCV usage for short, medium, and long

lanes in the FTA datasets. Substantial cost savings from LCV

use were possible in each lane distance category. However, at

distances above 1,000 mi (1,609 km), LCVs always lead to a

reduction in the total logistics cost when operating under

high GVW limits. Only turnpike doubles running under low GVW

limits at freight rates higher than current single trailer

rates in some cases caused the total logistics cost to go up

at distances above 1,000 mi (1,609 km). In the short and

medium distance lanes, however, there was virtually no

correlation between shipping distance and percent change in

total logistics cost. Use of LCVs for short and medium

distances produced very large cost savings in some cases and

very large cost increases in others.

Annual Lane Ton-Mileage

Annual lane ton-mileage is a composite measure of annual

lane volume and lane distance. In the FTA datasets, values

ranged from 5,000 (7,295 metric ton-km) to

67,288,000 (98,173,192 metric ton-km). The average was

4,207,000 (6,138,013 metric ton-km) and the median was 736,000

(1,073,824 metric ton-km). Based on the observed effects of

annual lane volume and lane distance, higher values of annual

lane ton-mileage should increase the likelihood of higher cost

savings from the use of LCVs.

Table 13. Average percent reduction in total logistics cost

from use of LCVs for different traffic lane

distances.

Click HERE for graphic.

metric conversion 1 mi = 1.609 km

The average percent reduction in total logistics cost

resulting from the use of LCVs is shown in table 14 for three

levels of annual lane ton-mileage. The average impact of LCVs

increased markedly in going from one level to the next. The

jump in the average cost reduction percentages was especially

acute between the low and middle annual lane ton-mileage

categories. Most instances of LCVs causing the total logistics

cost to rise occurred in the under 250,000 annual ton-mi

(364,750 metric ton-km) category. Above 1,000,000 annual

ton-mi (1,459,000 metric ton-km), use of LCVs always resulted

in a decrease in total logistics cost except for turnpike

doubles operating under low GVW limits at LCV rates higher

than current single trailer rates. In the FTA datasets, nearly

44 percent of the cases involved annual flows greater than

1,000,000 ton-mi (1,459,000 metric ton-km), and nearly

three-fourths of the observations involved annual flows of

250,000 ton-mi (364,750 metric ton-km) or more.

Table 14. Average percent reduction in total logistics cost

from use of LCVs for different levels of annual

lane ton-mileage.

Click HERE for graphic.

Metric conversion: 1 ton-mi = 1.459 metric ton-km

The correlation between annual lane ton-mileage and the

percent reduction in total logistics cost resulting from the

use of LCVs is shown in table 15. Again the magnitudes of the

coefficients were rather low, although they were somewhat

better than the correlation coefficients found for product

value. Unlike the case for annual lane volume, the correlation

between annual lane ton-mileage and percent cost reduction was

consistently positive. In addition, the coefficients were

higher for the high weight capacity scenario than for the low

weight capacity scenario, particularly for turnpike doubles.

This again points out the lower efficiency of LCVs operating

under low GVW limits. As was the case for annual lane volume

and lane distance, annual lane ton-mileage appeared to be

better at indicating the likelihood of cost savings from LCV

usage than it was at predicting how much savings could be

expected.

Product Value and Annual Lane Volume

Neither product value nor annual lane volume showed much

correlation with the percent change in total logistics cost

resulting from use of LCVs, although the latter variable was a

good indicator of whether or not LCVs would be beneficial to a

shipper. These two variables were stratified and

cross-tabulated to analyze their combined effect. Three

categories of product value were defined:

Table 15. Correlation between annual lane ton-mileage and the

percent reduction in total logistics cost resulting from LCV

usage.

Click HERE for graphic.

. Low - less than $1.00 per lb ($2.20 per kg).

. Medium - between $1.00 per lb ($2.20 per kg) and $2.99 per

lb ($6.59 per kg).

. High - $3.00 per lb ($6.61per kg) or more.

Approximately one-third of the 176 observations in each FTA

dataset fell in each of the above categories. Annual lane

volume was also separated into three categories, defined as

follows:

. Low - less than 15,000 cwt (680,385 kg).

. Medium - between 15,000 cwt (680,385 kg) and 49,999 cwt

(2,267,905 kg).

. High - 50,000 cwt (2,267,950 kg) or more.

Again, approximately one-third of the FTA cases fell in each

of these categories. The cross-tabulation of product value and

annual lane volume therefore yielded nine combinations or

cells of these two variables with each cell having roughly the

same number of observations.

The results of the cross-tabulation analysis are

summarized in tables 16 through 19. Tables 16 and 17 cover

Rocky Mountain doubles under the low and high weight capacity

scenarios, respectively. Tables 18 and 19 cover turnpike

doubles.

Table 16. Combined effect of annual lane volume and product

value on total logistics cost using Rocky Mountain

doubles under existing GVW limits.

Click HERE for Graphic

Table 17. Combined effect of annual lane volume and product

value on total logistics cost using Rocky Mountain

doubles under higher GVW limits.

Click HERE for Graphic

Table 18. Combined effect of annual lane volume and product

value on total logistics cost using turnpike doubles

under existing GVW limits.

Click HERE for Graphic

Table 19. Combined effect of annual lane volume and product

value on total logistics cost using turnpike

doubles under higher GVW limits.

Click HERE for Graphic

The effect of product value was noticeable mainly for annual

lane volumes under 15,000 cwt (680,385 kg). At these low

volumes, the average and median percent reduction in total

logistics cost from using LCVs dropped considerably as product

value increased. In addition, the percentage of cases in which

LCV use resulted in an increase in the total logistics cost rose

significantly with each successively higher level of product

value when the annual lane volume was low. On the other hand,

even when the annual lane volume was low, the average percent

cost reduction was above the overall average when the product

value was low, as a comparison with table 7 showed.

Consequently, the likelihood that LCVs would actually increase

the total logistics cost appeared to be a consequence of the

combined effect of low lane volumes coupled with medium or high

product values.

For annual lane volumes in the medium and high range, the

effect of product value was not nearly as detectable. At these

volume levels, the average and median cost reduction percentages

were generally quite high regardless of the level of product

value. Under the high weight capacity scenario, however, the

average percent cost reduction for medium and high annual lane volumes

was always lowest for product values in the highest category.

Thus, high product value did appear to reduce the cost-

effectiveness of LCVs at medium and high annual lane volumes,

but generally not to the point where LCVs became

counterproductive.

Product Value and Annual Lane Ton-Mileage

The same categories of product value defined above were also

cross-tabulated with annual lane ton-mileage. The following

three levels of annual lane ton-mileage were used, based on a

plot of this variable against percent cost reduction as well as

the distribution of annual lane ton-mileage values in the FTA

datasets:

. Low - 350,000 ton-mi (510,650 metric ton-km) or less.

. Medium - Over 350,000 ton-mi (510,650 metric ton-km) but

less than or equal to 1,000,000 ton-mi (1,459,000 metric

ton-km).

. High - Over 1,000,000 ton-mi (1,459,000 metric ton-km).

The results of this cross-tabulation, summarized in tables 20

through 23, were similar to those obtained in the analysis of

product value and annual lane volume. When the annual lane

ton-mileage was low, the cost-effectiveness of LCVs was reduced

considerably as product value increased. The combination of low

ton-mileage and either medium- or high-valued products accounted

for the majority of cases in which LCV usage resulted in an

increase rather than a decrease in the total logistics cost.

When annual lane ton-mileage was low but product value was also

low, the average percent cost reduction was at or above the

overall average. Above 350,000 annual ton-mi (510,650 metric

ton-km), the effect of product value was not evident except

under the high weight capacity scenario. At medium

and high levels of annual lane ton-mileage, high product

values had a noticeable dampening effect on the

cost-effectiveness of LCVs operating under higher GVW limits.

However, the effect was not strong enough to make LCVs

counter-productive. In fact, the cost reduction percentages

were still quite large and in some cases were still well above

the overall average.

INTRAMODAL DIVERSION TO LCVs

The FTA identified the truck configuration which had the

lowest total logistics cost for each product-lane observation.

As the results presented in the previous sections have shown,

in most cases the lowest cost configuration was some type of

LCV. Whether shippers in these cases would actually switch to

LCVs is a question which cannot be answered directly from the

responses to the shipper survey or from the FTA output. The

answer undoubtedly depends on many complex factors, and each

shipper may weight each factor differently. Nevertheless,

total logistics cost is presumably one of the more important

considerations .

The simplest approach to estimating an intramodal LCV

diversion rate would be to assume that a shipper will use

whichever trailer configuration produces the lowest total

logistics cost. Most shippers, however, would probably not

switch to an LCV configuration unless the resultant cost

savings exceeded some fixed percentage of the firm's current

total logistics cost. What that percentage might be is

difficult to determine. Therefore, the decision was made to

estimate a range of diversion rates based on different cost

savings thresholds.

Table 24 indicates the sensitivity of intramodal LCV

diversion rates to several factors: GVW limits, minimum

percent reduction in total logistics cost needed for diversion

to occur, and LCV freight rates relative to current single

trailer rates. The table shows the percentages of FTA cases in

which diversion was assumed to occur as well as the

corresponding percentages of annual lane ton-mileage involved.

The following are some general observations from table 24:

. LCV diversion rates under the high weight capacity

scenario tended to be larger than those under the low weight

capacity scenario, although for minimum required cost

savings below 20 percent, the differences were fairly small.

. Not surprisingly, as the minimum percent cost savings

required to induce diversion increased, the diversion rate

dropped considerably, especially under the low GVW scenario.

Table 20. Combined effect of annual lane ton-mileage and

product value on total logistics cost using Rocky

Mountain doubles under existing GVW limits.

Click HERE for graphic.

Table 21. Combined effect of annual lane ton-mileage and

product value on total logistics cost using Rocky

Mountain doubles under higher GVW limits.

Click HERE for graphic.

Table 22. Combined effect of annual lane ton-mileage and

product value on total logistics cost using

turnpike doubles under existing GVW limits.

Click HERE for graphic.

Table 23. Combined effect of annual lane ton-mileage and

product value on total logistics cost using

turnpike doubles under higher GVW limits.

Click HERE for graphic.

Table 24. Percent of FTA cases and ton-mileage assumed to

divert to LCVs under different GVW limits, cost

savings thresholds, and relative LCV freight

rates.

Click HERE for graphic.

. LCV diversion rates were generally lower when LCV freight

harges were higher than current single trailer freight

charges. The differences between the diversion rates for

the two LCV rate levels were relatively small when the

minimum required cost savings was less than 10 percent.

For cost savings of 10 percent and higher, however,the

differences became increasingly larger.

. Regardless of the GVW scenario or the level of LCV

freight charges, the estimated LCV diversion rates in

table 24 were considerably high. Even with a minimum

required cost savings of 15 percent, the estimated

diversion rates were between 70 percent and 81 percent of

the cases in the FTA datasets. This corresponded to 82.5

percent to 96.9 percent of the total annual ton-mileage

for all of the survey observations.

4.EFFECTS OF LCV USAGE ON THE LOGISTICS COSTS OF

RAIL AND INTERMODAL SHIPPERS

INTRODUCTION

This chapter presents the FTA results for those

product-lane observations in which the current primary mode of

transportation was either rail boxcar or truck-rail

intermodal. The FTA datasets contained only 24 rail and 27

intermodal observations. Although these samples were too small

for any multivariate analysis, some simple tabulations of the

FTA results were made in order to gain some insight into the

possible impact of LCVs on the total logistics cost of rail

and intermodal shippers.

RAIL BOXCAR SHIPPERS

Table 25 summarizes the overall effect of LCV usage on

the total logistics cost of the rail boxcar shippers. If LCVs

were required to operate under 80,000-lb (36,288-kg) GVW

limits, two-thirds of the rail shippers in the sample would

incur a higher total logistics cost in the traffic lane as a

result of switching to Rocky Mountain doubles. The average

cost increase would be 36 percent, and in half of the cases,

the cost increase would be 14 percent or more. The

consequences of switching to turnpike doubles under existing

GVW limits would be even more severe. The cost increase would

be at least 26 percent in half the cases and would average

nearly 49 percent. On the other hand, the FTA predicted a much

different outcome for the high weight capacity scenario. If

Rocky Mountain doubles were allowed to operate at gross

vehicle weights up to 115,000 lb (52,164 kg), most of the rail

boxcar shippers would experience a decrease in total logistics

cost as a result of switching to the LCV mode. Although the

average effect would be a 3.5 percent increase in the total

logistics cost, in half of the cases, the total logistics cost

would decrease by at least 13.7 percent. For turnpike doubles

operating at gross vehicle weights up to 131,000 lb (59,422

kg), the FTA results were even more favorable for LCVs. In

roughly 7 out of every 10 cases, switching from rail boxcars

to turnpike doubles resulted in a lower total logistics cost.

The average cost reduction was around 10.5 percent, and in

half of the cases, the reduction was 25 percent or more.

Table 26 indicates that rail boxcars would continue to

be the most cost-effective means of transport for current rail

users under the low GVW scenario, while turnpike doubles would

be the cost-effective mode in most cases under the high weight

capacity scenario.

Under both GVW scenarios, the variation in the FTA

results was extremely large. As table 25 shows, for some rail

shippers, switching to LCVs could result in cost increases of

well over 200 percent under the low GVW limit scenario and

over 100 percent in the high

GVW limit situation. At the other extreme, switching to LCVs

could reduce some rail shippers' total logistics cost by well

over 50 percent.

Table 25. Overall effect of LCVs on the total logistics cost

of rail boxcar shippers.

Click HERE for graphic.

Notes: Negative percentages imply an increase in total

logistics cost. Percentages in the rightmost column

are based on a sample size of 24 cases. The FTA used

an LCV freight rate of $1.30 per mi ($0.81 per km).

Table 26. Transportation mode with the lowest total

logistics cost for rail boxcar shippers.

Click HERE for graphic.

Note:Percentages are based on a sample size of 27 cases.

To understand better why the overall effects of LCVs on

rail boxcar shippers was so wide ranging, the shippers'

current rail freight charge per mile was plotted against the

percent reduction in total logistics cost. Figures 3 and 4

display the results for Rocky Mountain doubles under the low

and high weight capacity scenarios, respectively. The

corresponding charts for turnpike doubles are shown in figures

5 and 6.

Click HERE for graphic.

Each figure shows a distinct relationship between the

rail freight charge and the percent change in total logistics

cost resulting from a switch to LCVs. The correlation between

the two variables, shown in table 27, was fairly strong. The

relationship, however, was clearly nonlinear. In fact, below a

rail charge of roughly $2.50 per mi ($1.55 per km), each

additional small decrease in the rail rate tended to result in

an increasingly larger percent increase in total logistics

cost from switching to an LCV. In other words, below a rail

rate of around $2.50 per mi ($1.55 per km), the relative

effect on total logistics cost of switching from rail boxcars

to LCVs was much more sensitive to changes in the rail freight

rate, implying much higher cross-elasticities in that range.

This suggests that if a rail carrier's current freight charge

was at or below this threshold rate, the carrier might have to

reduce its rate by only a very small amount in order to offset

any logistics cost benefits of LCVs and to prevent its traffic

from diverting.

Table 27. Correlation between rail freight charge per mile

and percent reduction in total logistics cost

from switching to LCVs.

Click HERE for graphic.

The rail freight rate at which there would be no

difference in the total logistics costs of rail boxcars and

LCVs varied by type of LCV and GVW limits. For Rocky Mountain

doubles, this rate was between $2.50 and $3.50 per mi ($1.55

to $2.18 per km) under the low weight capacity scenario and

between $2.00 and $2.50 per mi ($1.24 to $1.55 per km) under

the high weight capacity scenario. For turnpike doubles, the

breakeven rail rate was between $2.50 and $4.50 per mi ($1.55

to $2.80 per km) under existing GVW limits and around $2.00

per mi ($1.24 per km) under higher allowable GVW limits.

INTERMODAL SHIPPERS

The overall effects of switching from TOFC/COFC

intermodal transportation to LCVs on the total logistics cost

of intermodal shippers are summarized in table 28. As was the

case for rail boxcar shipments, LCVs were generally less

cost-effective compared to intermodal service when operating

under restrictive GVW limits than when operating under much

higher limits. In a slight majority of cases, use of Rocky

Mountain doubles under the

low weight capacity scenario entailed an increase in the total

logistics cost. The average cost increase was 4.2 percent. For

turnpike doubles operating under existing GVW limits, nearly

half of the cases showed an increase in the total logistics

cost and half showed a decrease. On the average, however, the

effect of switching to turnpike doubles was about a S percent

increase in the total logistics cost. Under the high weight

capacity scenario with its more liberal GVW restrictions, both

types of LCV had much more favorable impacts. Use of Rocky

Mountain doubles reduced the total logistics cost in

two-thirds of the cases. The average cost reduction was 6.6

percent and in half of the cases the reduction was at least

11.4 percent. Use of turnpike doubles at higher GVW limits

reduced the total logistics cost in 8 out of every 9 cases.

The average cost reduction was over 16 percent and in half of

the cases the reduction was over 26 percent. Table 29

indicates that none of the three modes was predominant under

the low weight capacity scenario, while turnpike doubles were

the most cost-effective mode in nearly 9 out of 10 cases under

the high weight capacity scenario.

The variation in the effect of LCVs on the total

logistics cost of intermodal shippers was not nearly as wide

ranging as was the case for rail boxcar users. In fact, the

range in the cost reduction percentages was about the same as

that for single trailer truckload shippers. In an attempt to

explain the variation, intermodal freight rates were plotted

against percent reduction in total logistics cost. The results

are displayed in figures 7 through 10 for each combination of

LCV type and GVW scenario.

Table 28. Overall effect of LCVs on the total logistics cost

of intermodal shippers.

Click HERE for graphic.

Notes: Negative percentages imply an increase on total

logistics cost. Percentages in the rightmost column are

based on a sample size of 27 cases. The FTA used an LCV

freight rate of $1.30 per mi ($0.81 per km).

Table 29. Transportation mode with the lowest total logistics

cost for intermodal shippers.

Click HERE for graphic.

Note: Percentages are based on a sample size of 27 cases.

The relationship between intermodal freight rate and LCV

percent cost reduction was not nearly as well defined. Table

30 shows the correlation between the two variables. Except for

Rocky Mountain doubles under the low weight capacity scenario,

the correlation coefficients were not as large as those shown

in table 27 between rail freight rate and LCV percent cost

reduction. Nevertheless, some relationship was discernible.

Under the low weight capacity scenario, switching to LCVs

tended to increase the total logistics cost when the

intermodal freight rate was below average and to decrease the

total logistics cost when the intermodal freight rate was

above average. With higher GVW limits, the FTA results showed

that most intermodal shippers would have a lower total

logistics cost using LCVs. Those few shippers who would have a

higher total logistics cost were currently paying below

average intermodal freight rates. Given the fairly narrow

range of intermodal freight rates in the FTA datasets, the

effect of Rocky Mountain doubles was quite sensitive to

changes in intermodal freight rate under both GVW scenarios.

The effect of turnpike doubles was likewise highly sensitive

to changes in the intermodal freight rate, but only under the

low weight capacity scenario.

Table 30. Correlation between intermodal freight charge per

mile and percent reduction in total logistics cost

using LCVs.

Click HERE for graphic.

5. ESTIMATION OF NATIONWIDE LCV USAGE

INTRODUCTION

The second major objective of this study was to apply

the results of the FTA program to the difficult problem of

estimating the nationwide use of LCVs. The original intent was

to predict both intramodal as well as intermodal diversion.

However, the number of rail boxcar and truck-rail intermodal

traffic lane observations in the shipper survey dataset was

too small to develop a credible model of rail-to-LCV or

intermodal-to-LCV diversion. Consequently, the effort was

restricted to estimating the extent of diversion from single

trailer truckload shipments to LCVs.

This chapter begins by presenting a simple intramodal

diversion model based on the results of the FTA. It then

describes how this model was applied to data from the 1987

Truck Inventory and Use Survey (TIUS) to estimate the

nationwide diversion of single trailer truckload shipments to

LCVs. The resulting projections are then presented and

compared with projections from previous research.

INTRAMODAL DIVERSION MODEL

The FTA results indicated that annual lane ton-mileage

was an important factor in determining whether or not a

truckload shipper would benefit from LCVs. Product value was

also a factor, but its effect was mainly visible at only very

low annual lane volumes. Therefore, a simple classification

model based on annual lane ton-mileage was developed using the

shipper survey data and FTA results.

. The model incorporated the following four factors:

. Annual lane ton-mileage.

. Gross vehicle weight limits.

. Minimum percent cost savings required before truckload

shippers will be induced to switch from single trailers to

LCVs.

. LCV freight rates relative to current rates for single

trailer truckload shipments.

After careful study of plots of annual lane ton-mileage versus

LCV percent cost reduction, the following four levels of

annual lane ton-mileage were chosen for the model:

. Less than 50,000 (72,950 metric ton-km).

. 50,000 or more but less than 250,000 (364,750 metric

ton-km).

. 250,000 or more but less than 1,000,000 (1,459,000).

1,000,000 or more.

The gross vehicle weight limit factor was represented by the

high and low weight capacity scenarios. Because the minimum

cost savings necessary to induce diversion to LCVs could not

be deduced from the shipper survey and probably varies among

shippers anyway, the model included levels of 0, 5, 10, 15,

20, and 25 percent. Finally, the relative LCV freight rate

factor was represented by the following two levels:

. Same - shippers will pay the same rate for LCV service

that they currently pay for single trailer truckload

transport.

. Higher - shippers will pay a somewhat higher rate for

LCV service, reflecting the generally higher operating

cost per mile of LCVs.

The model was essentially a four-dimensional matrix

whose cells were formed by the various combinations of the

above four factors. The contents of each cell specified the

percentage of annual ton-mileage estimated to divert to LCVs

under the particular combination of factors defining the cell.

These percentages were taken from the results of the FTA as it

was applied to traffic lane data from the shipper survey.

The model is presented in table 31. The percent of

ton-miles diverting increases as the level of annual

ton-mileage increases. The model assumes that no diversion

will take place for annual flows below 50,000 ton-mi (72,950

metric ton-km) in a traffic lane. Flows in that range were

judged to be too small to justify use of LCVs even though the

FTA results indicated that, in a few such cases, LCVs would

reduce the total logistics cost. A company which ships

48,000-lb (21,773-kg) payloads in a 200-mi (322-km) traffic

lane 10 times a year has an annual freight flow of 48,000

ton-mi (70,050 metric ton-km) in the lane. That is a rate of

less than one truckload shipment per month. The assumption was

made that a company would not consolidate such a small number

of annual shipments by switching to LCVs.

ESTIMATION PROCEDURE

Correct application of the intramodal diversion model

requires nationwide statistics on individual shipper or

company traffic flows or a representative sample of company

shipments. Unfortunately, such data were not available when

this study was conducted. As a convenient but less than ideal

substitute, individual truck data from the 1987 Truck

Inventory and Use Survey (TIUS) were used.

Table 31. Percent of annual traffic lane ton-mileage that would

divert to LCVs.

Click HERE for graphic.

The TIUS is conducted every five years by the U.S. Bureau of

the Census. It is based on a stratified random sample of

private and commercial trucks registered or licensed in each

State during the survey year. The sample is divided into the

following five strata:

. Stratum 1 - pickup.

. Stratum 2 - panel truck, van, utility vehicle, jeep, and

station wagon on truck chassis.

. Stratum 3 - small single-unit truck with gross vehicle

weight rating (GVWR) less than 26,000 lb (11,794 kg).

. Stratum 4 - large single-unit truck with GVWR greater

than or equal to 26,000 lb (11,794 kg).

. Stratum 5 - truck tractor.(7)

Only stratum 5 was needed to estimate nationwide LCV usage.

This stratum contained 34,619 individual truck records.

TIUS sample data are normally used to generate national

estimates of the number of trucks and truck miles for various

truck characteristics such as truck type and axle arrangement,

major use, range of operation, operating weight, overall

length, type of operation and jurisdiction, kind of carrier,

and products carried. By making some assumptions, it is also

possible to derive estimates of ton-mileage from TIUS sample

data.

The TIUS provides a wide variety of data on truck

activity, but it does have a few limitations, particularly

with regard to the use to which it was put in this study. A

major drawback is the fact that TIUS records represent

individual truck activity rather than individual shipper

activity. Another limitation is the absence of data on

individual shipments. Thus, it is not possible to determine

origin-destination or traffic lane flows directly from the

TIUS data. A third weakness is the fact that reported annual

mileage in many cases is based on the truck owner's own rough

estimates rather than on actual odometer readings.(8)

Despite these problems, the TIUS dataset does contain

several data items that appeared to be especially useful for

estimating nationwide LCV usage. Table 32 lists the pertinent

fields or data items that were selected from the TIUS records

on the public use tape.

The remainder of this section describes the procedure

that was followed to derive estimates of intramodal diversion

to LCVs nationwide.

Table 32.Definitions of relevant data items selected from 1987

TIUS public use records.

Click HERE for graphic.

In the first step, TIUS stratum 5 records meeting the

following criteria were selected:

. Vehicle type (VEHTYP) is truck tractor pulling trailer(s).

. Body type (BODTYP) is basic enclosed dry cargo van.

. Single trailer configuration with at least two axles on the

trailer (AXLRE codes 14, 16, 18, 19, 20, and 21).

Records in TIUS stratum 5 include tankers, grain hoppers, dump

trucks, refrigerated vans, wreckers, beverage trucks, straight

trucks, as well as dry vans. Because the shipper survey data

included only truck shipments made in enclosed dry cargo

trailers, only those TIUS records involving truck tractors

pulling such trailers were considered. Single-axle semi-

trailers were excluded because they are generally small

trailers, usually 28-ft (8.5-m) in length. The assumption was

made that any shipper who currently uses small trailers is

more likely to switch to a larger sized single trailer before

switching to an LCV. The TIUS dataset also contains records

for tractors pulling two and even three trailers. These

records were also excluded because the shipper survey data did

not include any lane observations involving double and triple

trailer configurations.

In the second step, each selected vehicle's

weight-limited payload was computed using the following

formula:

Weight-limited payload (lb) = MAXWGT - EMPWGT (1)

In the third step, each selected vehicle's cube-limited

payload weight was computed as two-thirds of the

weight-limited payload weight. Analysis of the shipper survey

data revealed that the average payload of cube-limited

shipments was two-thirds the average payload of weight-limited

shipments (the actual factor was 0.6615).

The fourth step involved estimating the vehicle's annual

weight-limited ton-mileage for trips over 200 mi (322 km) from

its home base. The following equation was used:

Annual weight-limited ton-miles=ANNMIL x ( PLONG / 100 )

x (PCARWT / 100 )

x ( weight-limited payload . 2,000 ) (2)

The vehicle's annual cube-limited ton-mileage for trips

over 200 mi (322 km) from its home base was estimated in step

5 from the following formula:

Annual cube-limited ton-miles = ANNMIL x ( PLONG . 100 )

x (PCARSZ . 100 )

x ( cube-limited payload . 2,000 ) (3)

In step 6, the vehicle's total annual weight-limited and

cube-limited ton-mileage were each allocated to each of the

following commodity groups according to the percent of the

vehicle's annual mileage accounted for while carrying the

particular commodity:

. Processed foods and tobacco products (SIC 20 & 21) -PRFOOD.

. Textiles and apparels (SIC 22 & 23) - TEXTIL.

. Lumber and fabricated wood products (SIC 24) - LUMBER.

. Furniture and/or hardware (SIC 25) - FURN.

. Paper and paper products (SIC 26) - PAPER.

. Chemicals and/or drugs (SIC 28) - CHEM.

. Petroleum and petroleum products (SIC 29) - PETROL.

. Plastics and/or rubber products (SIC 30) - PLASTK.

. Building materials, such as gravel, sand, concrete, flat

glass, etc. (SIC 32) - BLDGMA

. Glass products (SIC 32) - GLASS.

. Fabricated metal products (SIC 34) - FABMTL.

. Machinery, including electrical or nonelectrical and

electronic (SIC 35 & 36) - MACHINE.

. Miscellaneous products of manufacturing (SIC 38 & 39) -

MSCMFG.

Each of these commodity groups was represented by at least

four observations in the shipper survey data. The commodity

groups that were not included consisted of the following: live

animals, fresh farm products, unrefined mining products, logs

and forest products, primary metal products, transportation

equipment, refuse, industrial water, and hazardous wastes.

Another TIUS commodity group excluded from consideration was

mixed cargo, which includes small packages. This group was

assumed to represent all less-than-truckload (LTL) shipments.

The intramodal diversion model was applied in the

seventh step. The ton-mileage diversion percentages shown in

table 31 were used to estimate how much of each vehicle's

commodity-specific annual weight-limited and cube-limited

ton-mileage would be diverted to LCVs under the different GVW

scenarios, LCV freight rate assumptions, and minimum

percent cost savings thresholds. The results for each vehicle

were then multiplied by the vehicle's expansion factor

(EXPFAC) to obtain nationwide estimates for the entire truck

population.

In the eighth and final step, the estimates of

nationwide diverted ton-mileage were converted to

vehicle-miles. This was accomplished by dividing the

weight-limited tonmileage estimates by the average

weight-limited payload and the cube-limited ton-mileage

estimates by the average cube-limited payload. From the TIUS

data the average weightlimited payload was determined to be

47,756 lb (21,681 kg), which was rounded up to 48,000 lb

(21,773 kg), and the average cube-limited payload was found to

be 31,837 lb (14.441 kg). which was rounded up to 32,000 lb

(14,515 kg).

An important assumption in the procedure described above

was that, when a truck is neither cube-limited nor

weight-limited, there is no incentive to utilize LCVs. Thus,

the two key TIUS data fields were PCARSZ, the percent of

annual mileage the vehicle carried payloads that filled its

maximum cargo size, and PCARWT, the percent of annual mileage

the vehicle carried payloads that weighed the maximum cargo

weight. The existence of these two data items was the main

reason for choosing to use the TIUS dataset despite its

serious limitations. Close inspection of the TIUS data,

however, revealed that these two percentages were often

overlapping. For example, a truck owner may have reported a

cube-limited percentage of 75 and a weight-limited percentage

of 60. Thus, for 75 percent of its annual mileage the truck

was running cube-limited and for 60 percent of its annual

mileage it was running weight-limited. Clearly there must have

been some overlapping mileage in which the vehicle was

operating both cube-limited and weight-limited. There was no

easy resolution of this problem except to compute and report

diverted cube-limited and weight-limited ton-mileage

separately. Each of these estimates by itself understates the

total LCV ton-mileage, while the sum of the two overstates it.

ESTIMATED INTRAMODAL DIVERSION

Tables 33 through 36 present the results of the

intramodal diversion estimation procedure described in the

previous section. The high estimates in these tables are the

amount of diversion that would occur if shippers switched to

LCVs for any amount of reduction in total logistics cost. The

low estimates are the amount of diversion that would occur if

shippers switched to LCVs only if the resulting reduction in

total logistics cost was at least 25 percent. All the

estimates shown in the tables are for 1987. To obtain

estimates of diverted vehicle-miles for 1992, the numbers in

the tables can be multiplied by a factor representing the

growth in truck travel between those two years. Published

highway statistics show that the total number of rural

truck-miles of travel by tractor-trailer combinations rose

from 55,978 million (90,069 million truck-km) in 1987 to

63,984 million (102,950 million truck-km) in 1992.~9 ' ' This

was an increase of 14 percent. Hence, an appropriate growth

factor to apply to the numbers in tables 33 through 36 would

be 1.14 to obtain 1992 estimates.

Table 33. Estimated truck vehicle-miles (in millions)

diverting to LCVs under existing GVW limits

with LCV freight rates same as current single

trailer truckload rates.

Click HERE for graphic.

Table 34. Estimated truck vehicle-miles (in millions)

diverting to LCVs under existing GVW limits with

LCV freight rates higher than current single

trailer truckload rates.

Click HERE for graphic.

Table 35. Estimated truck vehicle-miles (in millions)

diverting to LCVs under higher GVW limits with LCV

freight rates same as current single trailer

truckload rates.

Click HERE for graphic.

Table 36. Estimated truck vehicle-miles (in millions)

diverting to LCVs under higher GVW limits with LCV

rates higher than current single trailer truckload

rates.

Click HERE for graphic.

The percentages appearing at the bottom of tables 33 to

36 were based on an estimated total truck-mileage of 4,063.2

million (6,537.7 million truck-km) in 1987. This total was

derived from the TIUS records selected in the diversion

estimation procedure. It represents the number of miles

operated by truck tractors pulling single enclosed dry van

semi-trailers with at least two axles while hauling one of the

13 selected commodity groups on trips at least 200 mi (322 km)

long. It accounted for about 22 percent of the 18,667.4

million long-range truck-mi (30,035.8 million truck-km) made

in 1987 for the 13 selected commodity groups for all body

types and combinations of tractor-trailers."" In turn, the

total long-range truck-mileage for the 13 commodity groups

comprised about 49 percent of the 38,383 million long-range

truck-mi (61,758.2 million truck-km) made in 1987 for all

truck-tractors hauling any kind of commodity, including mixed

cargo, in any kind of body type and configuration of

trailers.(11)

The estimates of nationwide intramodal diversion to LCVs

ranged from 0.5 billion to 1.5 billion truck-mi (0.8 billion

to 2.4 billion truck-km) for weight-limited freight and from

0.75 billion to 2.1 billion truck-mi (1.2 billion to 3.4

billion truck-km) for cube-limited freight. As mentioned

earlier in this chapter, there was an undetermined amount of

overlap between the weight-limited and cube-limited truck

mileage so that adding the two estimates would likely

overstate the amount of diversion.

The broad range of estimated nationwide intramodal

diversion indicates the significant extent to which such

factors as GVW limits, LCV freight rates relative to single

trailer truckload rates, and minimum required cost savings can

affect the amount of diversion that might occur. For example,

going from existing GVW limits to higher limits more suited to

LCVs could divert an additional 288.8 million truck-mi (464.7

million truck-km) if LCV freight charges were the same as

single trailer truckload rates. The significance of how much

reduction in total logistics cost would be required to induce

shippers to use LCVs is reflected in the differences between

the high and low estimates, which ranged from 329.7 million

truck-mi (530.5 million truck-km) to 1,285.9 million truck-mi

(2,069.0 million truck-km), depending on the GVW limit and LCV

freight rate scenario. Whether or not the estimates are

realistic is impossible to determine, but they do show a great

deal of sensitivity to factors which are difficult to

determine.

COMPARISON WITH OTHER ESTIMATES

One way to gauge the reasonableness of the above

estimates is to compare them with estimates of intramodal

diversion from other studies. Such comparisons are burdened

with their own inherent difficulties because of differences in

types of LCVs considered, assumptions about LCV operating

costs and productivity, assumed GVW limits, estimation

methodology, assumptions about shipper and carrier acceptance

of LCVs, sources and quality of data, and so on. In the end,

all projections of nationwide LCV use are purely speculative.

Consequently, the only comparison made was with estimates

developed in a recent Transportation Research Board (TRB)

study.(3)

The TRB study analyzed four prototype truck-trailer

configurations with lower axle weights but higher than

currently allowed gross weights in accordance with a proposal

advanced by former Federal Highway Administrator Francis C.

Turner. The four Turner prototypes considered in the study

consisted of a seven-axle tractor-semitrailer combination,

nine- and eleven-axle double trailer configurations, and a

nine-axle B-train double trailer combination. These

prototypes were analyzed in terms of their effects on

pavement wear, bridge costs, safety, traffic operations,

freight costs, productivity, and use.

The market for the double-trailer prototypes was

divided into a number of sectors defined by trailer body

type, private versus for-hire carrier, and local versus

intercity travel. The various body types included dry vans,

reefers, flatbeds, dry bulk trailers, and tankers. Dry vans

were further segmented into less-than-truckload and truckload

shipments and the latter were further divided into

low-density and high-density commodity groups. A market

potential of low, moderate, or high was assigned to each of

these sectors, based on a review of past projections of LCV

usage, estimates of the costs of the Turner prototypes,

interviews with carriers, and current use of multiple trailer

configurations. Each market rating was then quantified as

follows:

. A low market potential meant a 10 percent potential shift

of traffic to Turner prototypes.

. A moderate market potential corresponded to a potential 33

percent shift.

. A high market potential entailed a potential 67 percent

shift.

These diversion rates were then applied to current

vehicle-miles of travel by 5-or-more axle

tractor-semitrailers, S-axle truck-trailers, and 5- and

6-axle double trailers in each market sector. The total

number of vehicle-miles of travel for all combination trucks

was taken from the 1987 Highway Statistics publication. This

total was distributed among the various market segments based

on data from the 1982 TIUS, adjusted to account for the

increased use of twin trailer combinations between 1982 and

1987.

Of the many market sectors identified in the TRB study,

only the ones involving intercity truckload shipments in dry

vans closely matched the market sectors considered in the

current study. The TRB study assessed the market potential of

each of these sectors as follows:

. The private carrier, intercity, dry van, low-density

truckload market potential was rated as low to moderate.

. The for-hire carrier, intercity, dry van, low-density

truckload market potential was rated as low.

. The private carrier, intercity, dry van, high-density

truckload market potential was rated as moderate.

. The for-hire carrier, intercity, dry van, high-density

truckload market potential was rated as moderate.

The TRB study's estimates of traffic diversion to

Turner prototypes in each of the intercity dry van truckload

market sectors are shown in table 37. For comparison with

tables 33 through 36, low density freight may be regarded as

cube-limited and high density freight as weight-limited. In

comparing the TRB study's results with the estimates made in

the current study, it is more correct to use only tables 35

and 36 which cover the high weight capacity scenario.

Table 37. Estimated billions of truck miles diverting to

Turner prototypes in the intercity dry van

truckload market sectors.(3)

Click HERE for graphic.

Metric conversion: 1 truck-mi = 1.609 truck-km

The TRB study estimated between 0.7 and 1.4 billion

truck-mi (1.1 to 2.3 billion truck-km) would divert to Turner

prototypes in the low density intercity truckload freight

sectors. This overlapped the bottom of the range of 1.2 to

2.1 billion truck-mi (1.9 to 3.4 billion truck-km) of

diverted cube-limited freight estimated by the current study

under the high weight capacity scenario. On the other hand,

the TRB study's estimate of 4.5 billion truck-mi (7.2 billion

truck-km) of high density freight diverting to Turner

prototypes was three times higher than the highest estimate

of diverted weight-limited truck mileage in the current

study.

One reason for the large difference in the estimates of

diverted weight-limited freight was the fact that the TRB

study's estimates were based on a total truck mileage of 20.4

billion (32.8 billion truck-km) compared to 4.1 billion

truck-mi (6.6 billion truck-km) in the current study. Unlike

the current study, the TRB study considered not only

diversion from single trailer truckload shipments but twin

trailer shipments as well. The TRB study also did

not exclude any commodity groups. Consequently, a better

measure of comparison between the two studies is the estimated

market shares.

The TRB study estimated that 11 percent to 21 percent of

the total truck-miles of intercity low density dry van

truckload traffic would divert to LCVs, compared to an

estimated range of 31 to 51 percent in the current study. The

much lower market share in the TRB study reflects that study's

assessment of the market potential of the low density dry van

truckload sectors as low to moderate. For the high density

market sectors, whose market potential was rated as moderate,

the estimated market share was 33 percent, compared to a range

of 23 percent to 36 percent estimated by the current study.

Thus, a major difference between the results of the two

studies was due to the TRB study's low assessment of the LCV

market potential of low density or cube-limited truckload

freight traffic.

6. CONCLUSIONS

The underlying purpose of this study was to examine the

issue of LCV usage from the standpoint of the individual firm

or shipper rather than from the viewpoint of the motor carrier

or truck operator. This approach was quite different from that

of previous studies, which have often focused on the fixed and

variable costs of operating LCVs and the potential gains in

productivity which these truck configurations may offer to

the carrier. Based on these expected gains in carrier

productivity, previous studies have predicted significant

savings in freight costs for shippers. Freight costs, however,

comprise only a portion of the shipper's total logistics cost

which also includes the cost of carrying inventory. If by

switching from single trailer truckloads to LCVs, a shipper's

inventory carrying costs increase more than any savings in

transportation costs, then any productivity gains available to

the carrier will be lost to the shipper. The basic intent of

the current study was to determine the effects of LCVs on the

shipper's total logistics cost and how these effects might influence

the demand for LCV transport.

A major finding of this study was that, given sufficient

flows of a company's product in a lane, LCVs would generally

have a positive impact on the total logistics cost of firms

that currently ship in single trailer truckload quantities.

The actual amount that might be saved was difficult to predict,

but on the average, a reduction in total logistics cost between

13 percent and 32 percent could be expected, depending on the

type of LCV used, GVW limits, and the difference between LCV and

single trailer truckload freight charges. The higher the flow

and the longer the lane, the greater the likelihood that LCVs

would benefit the truckload shipper. In general, use of LCVs at

annual lane volumes below 5,000 cwt (226,795 kg) appears likely

to increase a firm’s total logistics cost, while at annual lane

volumes over 25,000 cwt (1,133,975 kg), LCVs are likely to

produce cost savings averaging between 16 Percent and 42

percent.

The influence of product value on the cost-effectiveness of

LCVs was surprisingly small. The correlation between product

value and the percent change in total logistics cost caused by

use of LCVs was exceedingly low. The effect of product value was

most apparent when the flow of a company's product in a lane

was below 15,000 cwt (680,385 kg) or 350,000 ton-mi (510,650

metric ton-km). In such cases, higher product values

significantly increased the likelihood that LCVs would greatly

elevate the company's total logistics cost. The main effect of

product value at moderate and large annual lane volumes was to

slightly reduce the average percentage of reduction in total

logistics cost resulting from the use of LCVs.

Predicting whether or not the use of LCVs for a given product

in a given lane will lower a company's total logistics cost

appears to be considerably easier than predicting the actual

magnitude of the change. Annual lane volume, lane distance, and

ane ton-mileage appear to be good indicators of whether or not

LCVs will be beneficial. However, none of these variables was

highly correlated with differences in total logistics cost

between

LCVs and single trailers. Perhaps an even better indicator is

the ratio of annual freight costs to annual inventory carrying

costs using single trailers in a lane. If the freight costs are

two or more times greater than the inventory carrying costs,

switching from single trailers to LCVs will in all likelihood

greatly reduce the total logistics cost. On the other hand, if

the inventory carrying costs are nearly the same as or greater

than the freight costs, then the chances are good that

switching from single trailers to LCVs will increase the

total logistics cost.

The effect of LCVs on the total logistics cost of rail

carload shippers was highly sensitive to the rail freight rate

per mile. While diversion from rail cars to Rocky Mountain

doubles does not appear likely, the results of the analysis

indicate that turnpike doubles when operating under higher than

existing GVW limits could reduce the total logistics cost

enough to induce some shippers to switch from rail to LCVs.

Because of the small number of rail observations in the shipper

survey data, it was not possible to estimate the amount of

diversion that might occur.

Turnpike doubles could also provide some serious competition

to rail-truck intermodal services if existing GVW limits were

to be raised. The analysis predicted a lower total logistics

cost for most of the intermodal shippers in the survey sample

as a result of switching to turnpike doubles.

Given the finding that LCVs are likely to reduce truckload

shippers' total logistics cost in most cases where traffic lane

volumes are moderate to heavy, the more difficult question still

to be answered is what effect will this have on the demand for

LCV freight services. Will shippers switch from single trailers

to LCVs in order to reduce their logistics costs? If so, how

much of a cost savings is required before a firm will make the

necessary adjustments in its production scheduling and

inventory management system to accommodate the higher shipment

quantities entailed by the use of LCVs? To what extent is the

demand for LCVs influenced by the potential savings in total

logistics cost? These are important questions which could not

be answered directly in this study. However, it was shown

that the level of cost savings necessary to induce diversion

from single trailers to LCVs has a tremendous impact on

projections of LCV demand.

With regard to the question about the relative importance

of logistics cost reduction to the demand for LCVs, a few

comments can be made. First of all, it became clear while

screening companies for eligibility to participate in the

shipper survey that many firms do not consider logistics

costs in general or inventory carrying costs in

particular when making modal choice decisions. In some cases,

particularly for small and even some medium-sized firms, it is

because the company did not have a logistics management system

sophisticated enough to determine what its inventory and other

logistics costs were. In other cases, the companies' product

costs and interest rates were so low that they no longer

considered inventory carrying costs in making daily

transportation decisions. Secondly, thereare other ways for

a company to reduce its total logistics cost than by

switching from single trailers to LCVs. A primary example of

this is thecurrent trend toward using the Just-in-Time concept

which trades off inventory and transportation costs by

maintaining small inventories, shipping in smaller quantities,

and using faster modes of transportation. Therefore, even though

this research has shown that LCVs are likely to have a positive

effect on shippers' total logistics cost, the trend toward

smaller shipment sizes and more rapid on-time deliveries would

seem to reduce the demand for LCVs from the standpoint of

many shippers.

REFERENCES

1. Providing Access for Large Trucks, Special Report 223,

Transportation Research Board, National Research Council,

Washington D.C., 1989.

2. Truck Weight Limits: Issues and Options, Special Report 225,

Transportation Research Board. National Research Council,

Washington DC, 1990.

3. New Trucks for Greater Productivity and Less Road Wear: An

Evaluation of the Turner Proposal, Special Report 227,

Transportation Research Board, National Research Council,

Washington, DC, 1990.

4. Evaluating Shipper-Related Productivity Gains to be Achieved

from Authorizing Increased Truck Lengths and Gross Vehicle

Weights, Center for Logistics Research, The Pennsylvania

State University, University Park, PA, December 1991.

5. Clyde Kenneth Walter, "Longer Combination Vehicles: Issues

and User Attributes,"Proceedings of the 35th Annual Meeting,

Transportation Research Forum, New York,NY, October 14-16,

1993.

6. The Effect of Size and Weight Limits on Truck Costs, Working

Paper, Jack Faucett Associates, Bethesda, MD, October 1991.

7. 1987 Census of Transportation, Truck Inventory and Use

Survey: Technical Documentation, U.S. Department of

Commerce, Bureau of the Census, Washington,DC, no date.

8. Patricia S. Hu, Tommy Wright, Shaw-Pin Miaou, Dennis J.Beal,

and Stacy C.Davis, Estimating Commercial Truck VMT of

Interstate Motor Carriers: Data Evaluation, Oak Ridge

National Laboratory, Oak Ridge, TN, November 1989.

9. Highway Statistics 1988, Table VM-1, Federal Highway

Administration, Washington,DC, 1989.

10. Highway Statistics 1992, Table VM-1, Federal Highway

Administration, Washington, DC. 1993.

11. 1987 Census of Transportation, Truck Inventory and Use

Survey: United States, U.S. Department of Commerce, Bureau

of the Census, Washington, DC, 1990.

ORNL-6840

INTERNAL DISTRIRUTION

1-5. M. S. Bronzini 17. D. E. Reichle

6. J. B. Cannon 18. R. B. Shelton

7. S. M. Chin 19. F. Southworth

8. D. L. Greene 20. Central Research Library

9. S. G. Hildebrand 21. Document Reference Section

10. P. S. Hu 22-23. Laboratory Records

11. M. A. Kuliasha 24. Laboratory Records - RC

12-16. D. P. Middendorf 25. ORNL Patent Office

EXTERNAL DISTRIBUTION

26. Dr. Douglas R. Bohi, Director, Energy and Natural

Resources Division, Resources for the Future, 1616 P

Street NW, Washington, DC 20036.

27. Dr. Thomas E. Drabek, Professor, Department of

Sociology, University of Denver, Denver, CO 80208-0209.

28-47. Arthur C. Jacoby, U.S. Department of Transportation,

Federal Highway Administration, Office of Policy

Development, Transportation Studies Division, Industry

and Economic Analysis Branch (HPP-11), Room 3324, 400

Seventh Street, SW, Washington, DC 20590

48. Mr. Calvin D. MacCracken, President, Calmac

Manufacturing Corporation, 101 West Sheffield Avenue,

Englewood, New Jersey 07631.

49. Ms. Jacqueline B. Shrago, Vice-Chancellor, Information

Technologies, Tennessee Board of Regnets, 1415

Murfreesboro Road, Suite 350, Nashville, TN 37217

50. Mr. George F. Sowers, Senior Vice President, Law

Companies Group, Inc., 114 Townpark Drive, Suite 250,

Kennesaw, GA 30144 5599.

51. Dr. C. Michael Walton, Paul D. and Betty Robertson Meed

Centennial, Professor and Chairman, Department of Civil

Engineering, College of Engineering, The University of

Texas at Austin, Cockrell Hall, Suite 4.2, Austin, TX

78712.

52. Office of Assistant Manager for Energy Research and

Development, DOE-ORO, P.O. Box 2001, Oak Ridge,

Tennessee 37831-8600.

53-54. Office of Scientific and Technical Information, U.S.

Department of Energy, P.O. Box 62, Oak Ridge, Tennessee

37831.