ECIV 790U: Intermodal Freight TransportClass Project Report

Estimation of Modal Diversion and Economic Benefits due to Rail Service Improvements in

South Carolina

ByOmor Sharif

Spring, 2013

IntroductionThe objective of this class project is to apply the intermodal transportation and inventory cost model (Highway to Rail Intermodal Version) known as ITIC-IM for modal diversion and economic benefit analysis. We review various aspects of the model- its purpose and use, the underlying methodology, data requirements and applications of the model. Also we show how to run the model and perform a sample analysis.

The Intermodal Transportation and Inventory Cost (ITIC) Model is a spreadsheet based computer model developed for performing policy analysis regarding long distance freight movement. It estimates the modal diversion or economic benefits in response to changes in transportation policy/infrastructure. The model replicates the decision-making process followed by a logistics manager who choose a mode and shipment size to supply his company’s inventory of a particular product. The alternative choices are explored for available modes and the exact cost is determined for each choice.

The Version 1.0 of the model was published in 2005 by US Department of Transportation and Federal Railroad Administration.

What is the purpose and use of the model?The ITIC Model gives estimate of mode diversion generated by a

change in the transportation levels of service or price. The change is typically caused by improvements in transportation infrastructure, transportation operations, or government policy. The model also provides policy analysis as it gives estimate of economic benefits associated with such changes.

The main application of the model is to estimate the diversion of highway freight traffic to rail intermodal service. However, the model can also be used to develop information needed for policy analysis of both rail to truck and truck to rail diversion. The model uses a single highway vehicle type—a conventional 5-axle tractor trailer combination—in a dry van configuration.

The model uses truck flow data inputs and can determine if rail intermodal can capture traffic from the common highway 5-axle tractor-trailers. For example, rail service improvements that lower logistics costs can make rail more efficient and this can be tested in the model.

How the model computes the mode diversion or economic benefits? To compute these estimates the model is employed as a disaggregate

demand model. One of the transportation alternatives is chosen based on minimum total logistics costs. This is then repeated for each of a large number of disaggregate observations from a representative sample of shipper movements. Statistics can then be computed on the resulting choices and the mode and shipment size shares determined for the sample as a whole, or for any sub-sample of interest.

The model needs an appropriate set of input data of typical freight movements prepared by the user. The input data must be entered in a predefined format in the spreadsheet as explained in user’s manual.

Objective

The objective of this project is to 1) Estimate the diversion of highway freight traffic to rail intermodal service 2) Estimate the economic benefits of changes 3) Understand the underlying methodology and formulas used for the estimates 4) Application of TRANSEARCH Database for analysis 5) Perform a case study on freight movements originating at South Carolina and destined to state of Florida and Virginia.

Contribution

The project has identified the suitable data sources for the model. Where data was not available inputs were developed for the model such as freight rates, intermodal rail miles, intermodal dray miles, number of railway interchanges that may be required for a movement.

Literature Review

The ITIC model is originated from transportation demand models performed at MIT starting in the 1970s and has undergone various developments since then. The model has been used in policy studies by both government and the private sector examining changes in infrastructure, transportation operations, pricing policy, government policy and possible advances in technology. The model documentation refers to many sources regarding previous research that has contributed the development of the model.

Model Features

The model has been used in many policy studies by government/private sectors examining changes in infrastructure, transportation operations,

pricing policy, government policy and possible advances in technology. In this section we will briefly review the economic theory which serves as the theoretical basis of the model, the diversion model itself, the components and organizational structure of model, the databases used as input, the processes which are used to prepare the data, and finally, the steps that are followed in running the model.

An Overview of the Model

The model consists of several components. Mode choice, shipment size, required level of service relationships produce the inputs needed for the model. It also requires some equipment characteristics tables and other tables to set the parameters of a particular scenario. The model is a discrete choice model that uses disaggregate freight movement databases.

Basic Nature of the ModelThe model uses disaggregate approach. It means that decision is made from the point of view of the individual decision maker (the individual shipper or receiver). It is important for freight than for passengers since the annual usage of products can range over several orders of magnitude. A product used at a rate of 10 pounds per year and a product used at a million pounds per year have a difference in use of 5 orders of magnitude. They require totally different treatment, have completely different logistics costs, and typically select different shipment sizes, and even different modes. Product value and shelf life are also extremely important to modal selection. High product value influences the amount of inventory that it is economical to hold, as does a short shelf life.

This transport costs are low per unit if a large shipment size is selected however the cost of carrying this product in inventory until it is used up is high. The transport cost is high for a smaller shipment sizes however has a lower inventory cost, but has more frequent reordering costs. An aggregate model does not consider the preferences/behavior of individual decision maker and results in an incorrect forecast. Fortunately it is possible to use the behavior of the individual decision maker as the economic rationale underlying the model.

Theoretical Basis

The theory of the firm is based on the assumption that each firm minimizes the costs required to produce a given quantity of output (maximizes profits). Economic theory treats transportation simply as one of the factors used in

production. However, transportation is different from other factors (such as land, labor, capital etc) in that it is not consumed directly, but is a service used only in processing other inputs or outputs. Also, it impacts the cost of other factors of production i.e. higher transport costs increases cost of inputs that require transport, which in turn, results in a higher cost for the delivered product.

The philosophy underlying the model is that the receiver is a rational economic decision maker who attempts to minimize the total cost of acquiring the inputs he needs for production, shipping them to the place he needs them in the process, storing them until their use and protecting the company against possible shortages during the process. In short, the receiver attempts to minimize total logistics costs for the delivered product. This involves not only the selection of the mode of transport to be used, but also the selection of the supplier of the product, the choice of inventory control system, the location of warehouses and the firm’s overall strategy for serving the market. The process is too complex to address in detail at this point, however, the basic theoretical foundation of the model is based on this concept.

Variables Affecting Choice of Supplier, Shipment Size, and Mode

The factors influencing a mode are complex choice are highly interdependent. They involve tradeoffs mainly between transportation cost, overall transit time and delivery reliability. Typically, the receiver is the buyer of goods, the shipper is the seller of the goods. The shipper is typically the receiver's "agent" in the process and it is his wishes that are honored in the size of shipment and the choice of mode. It is therefore appropriate to view the process as involving a single decision-maker—the shipper/receiver.

The most important tradeoffs involve the annual use of a product by the receiver. High annual use of a product allows the receiver to order large replacement shipments and to take advantage of the low transport costs afforded by economies of scale in shipping associated with large shipment sizes. High value of the product imposes a penalty to ordering more than can be readily used by tying up capital in inventory. Excess inventory can be avoided by ordering product more frequently in smaller shipment sizes. Small shipment sizes carry their own penalties. Ordering is a costly process. Smaller shipment sizes typically carry high unit cost of transportation, and if the shipment size is smaller than a full vehicle load, the load must be picked up at the origin by the freight carrier and consolidated before shipment, then

deconsolidated and delivered at the destination end. Most LTL, less than truckload, trucking, parcel carriers and airfreight systems perform consolidation/deconsolidation of smaller shipments into full vehicle loads. The consolidation and deconsolidation processes are also expensive, sometimes exceeding the cost of linehaul transportation.

Other variables can also play an important role. The density of a product influences the choice of vehicle either by loading "heavy," in which case payload is important, or loading "light," in which case cube is more important. Shelf life influences choice of mode by placing a premium on transit time, where longer travel time leads to less time available on the grocer's shelf before the product spoils. Loss and damage may lead to a need for emergency shipments. Many variables turn out to be important to the process.

The variables typically involved in the decision process are shown in the figure below. This figure shows that many of the variables interact with other variables to produce the conditions that result in the final choice of supplier, shipment size and mode. The results contribute to the shipper/receiver's total logistics cost per unit.

Tradeoffs made by the shipper/receiver

Most of these variables affecting the choices of the receiver have been incorporated into the ITIC Model. The program develops the tradeoffs that would be made by a receiver who is attempting to minimize the total logistics costs associated with maintaining an inventory of the product for use in manufacturing or wholesale trade. The variables are used to develop each of the individual cost factors listed on the right hand side of the figure above. They include the type of receiver, variables that describe the product, information on the current mode of transport and potential new modes and the attributes of the product being carried.

These variables are used to write equations for each of the components of the receiver's total logistics costs as a function of the principal choice variables (i.e. choice of supplier, choice of mode and choice of shipment size). Total logistics costs can be expressed in cost per unit,

cost per hundredweight or annual cost. Transport charges are added to logistics costs to give the total transportation and logistics cost of the strategy. If different suppliers are considered, with different purchase costs, the total delivered cost per unit or per hundredweight is given. Most receivers will select that strategy with the minimum total delivered cost. This program can be used to examine those circumstances under which one mode will be chosen over other modes.

Truck-to-truck diversion involves decisions made by carrier management as to what equipment to use to accomplish a particular movement. By contrast, rail-to-truck, or truck-to-rail diversion involves a decision by the shipper/receiver to use another entirely different mode of transport. This "between modes" type of decision is more complex, involving the evaluation of tradeoffs in equipment availability, transit time and reliability of delivery, freight loss and damage experience and the size of the potential shipment and its suitability for movement on the mode in question. The shipper’s rationale for making these decisions must be modeled if these tradeoffs are to be evaluated properly.

Cost of Movement to the Receiver

The receiver select a mode and shipment size and supplier such that it will minimize the total logistics cost of the goods being shipped to the receiver. Demand for a mode changes in response to changes in service or cost in the mode, which may impact the individual shippers' own business and the other alternatives available. However, the model assumes that all of the product used annually will move by one of the alternatives.

In key variables in the model may be grouped into three major groups: 1) Shipper/receiver attributes. 2) Commodity attributes. 3) Transport attributes.

Shipper/receiver attributes

The shipper/receiver characteristics is most important as noted before and the annual use of the product by the receiver. Rail is capable of handling larger individual shipments than truck. The typical carload can handle shipment weights up to 200,000 pounds, or more, while a maximum single unit truckload payload is around 50,000 pounds. Rail carload shipments of 100 tons are routine and multicar shipments of 1,200 tons or more can be handled on the same bill of lading. Unit trains moving as much as 10,000

tons (20 million pounds) are also common. By contrast, if a shipper must take a 200,000-pound shipment in order to use rail, instead of the 20,000-

pound shipment he would like to take, it could result in thousands of dollars of unwanted inventory cost. Shipper modal choice behavior, then, depends importantly on the amount of product used annually.

Commodity attributes

Commodity attributes are also important because the product type being shipped determines the loading and handling requirements as well as the maximum size of shipment that can be accommodated in a given piece of equipment. These variables include: density, value per pound, shelf life, typical packaging. Some of these data are available in a Commodity Attribute File from the Federal Railroad Administration. The relevant product data are appended to the individual movement observation in the input data prepared by the user for input into the model.

Transport attributes

Some transport attributes related to the mode can be important such as availability of equipment, transit time, reliability, loss and damage experience.

Variables Affecting Choice of Supplier, Shipment Size and Mode in Freight Transportation

Receiver Minimizes Total Costs / Unitwhich consists of:

Logistics Cost per Unitorder cost

load/unload costcapital carry in transitcapital carry in storage

storage costshelfloss in transitfiling L&D claims

capital carry on L&Dsafety stock carrying costemergency shipment cost

Total Logistics Costs per UnitTransport Charges

Trans & Logistics Cost per UnitPurchase Cost

Total Costs per Unit

Type of Receiver: Producer Wholesaler Retailer Government Individual consumer

Affects buying decisions: 1) Buy from original producer 2) Buy from wholesaler who performs consolidation, deconsolidation and inventory functions 3) Buy from retailer who buys from wholesaler

Type of product: Product annual use Value/lb. of product Shelf life of product Storage requirements

Affects size of shipment and ability of receiver to hold product in inventory

Commodity Attributes: Density of product Cube capacity of vehicle Weight capacity of vehicle

Affects loading of shipment by mode and possible need for consolidation of shipment with others

Transport level of service attributes: Transit time of mode Reliability of mode Waiting time for equipment

Affects cost of capital tied up in transit, safety stock holding cost and ability of mode to serve as emergency

Choice of shipment size affects: Cost of ordering Cost of loading and unloading Cost of pickup and delivery

Choice of Supplier

Choice of Mode

Choice of Shipment Size

Choice of mode affects: Cost of line haul transport

Choice of supplier affects: Length of haul Carriers available Purchase price of product

Total logistics cost

The above mentioned variables are incorporated into a "shipper's utility function” within the model. Models for estimating level of services attributes are included in the model. The total logistics cost is chosen as shipper's utility function. The cost is composed of ordering, transport, inventory and use of the product being shipped. Total logistics cost is the item that the shipper is attempting to minimize when he selects one mode of transportation over another or one shipment size over another. The components included in the shipper's total logistics cost function include: ordering cost, capital carrying cost in transit, capital carrying cost in inventory, warehousing cost, loading and unloading cost, safety stock carrying cost, cost of loss and damage claims.

Why ‘reliability’ is important?

The receiver restocks the inventory periodically so that inventory does not run out when needed. Reliability is the variability of lead-time of an order. Lead-time is the time passed between placing an order and actually receiving the shipment of product. Low reliability implies carrying more safety stock in the inventory to avoid running out. Amount of Safety stock considerably increases total logistics cost because it has to be carried in inventory continuously. Note that there is tradeoff between cost of carrying an extra safety stock and cost of running out of stock. There might also be a variability of the use rate of product by receiver which adds more uncertainty that current stock will last until new shipment is received.

Reliability impact the choice of mode and shipment size. For truck shipments the lead time is fairly predictable, however rail is not as reliable. The model can estimate the reliability of the modes to guarantee the required safety stock by some percentage (90%, 95% etc). It computes the standard deviation of transit time and based on that a safe reorder lead time is computed.

As explained in the user manual, ‘transit time reliability’ is the coefficient of variation of transit time (std. dev. divided by mean). Transit time is the ‘lead-time’ of a new order. Lead-time is the time spent between the time the order was placed by user and when the shipment was finally received by the user. Thus a higher value implies more variation in lead time and consequently lower value indicates better reliability.

The value is used to compute standard deviation of lead time (= coefficient of variation of lead time * mean lead time) for an order. Then std. dev. of user demand during lead time is computed as ‘User demand per day * standard deviation of lead time’. Now that we know mean and standard deviation of user demand during lead time we can construct the gamma distribution by finding alpha and beta. Reorder volume is then calculated using inverse gamma function is excel. The safety stock required can then be found. There is cost for carrying this safety stock which is a part of total annual cost of the chosen mode.

How the Model Can Be Used For Policy Analysis?

The model provides two analysis options. The first is the ‘Base Case’ in which the parameters are set to reflect current transportation levels of service and prices (BEFORE/current conditions). The second is the ‘Policy Case’ in which the parameters are set to reflect conditions that will exist after the new policies are placed into effect (AFTER conditions). The difference between the BEFORE and AFTER conditions is the IMPACT of the change in policy. A Policy Impact Report can be prepared by the user to summarize the findings. Note that the base case can also be used to validate the model parameters i.e. whether the results match with the real world data.

Data for Model

Population potentially impacted by the project or policy change

Project or policy to be tested

Representative sample of decision-makers drawn from the population

Disaggregate choice model used on each observation in the sample to deterrminne choices made by individual decision makerLogistics choices are: Where to acquire When to reorder What size shipment What mode to use

1) Modify the transport level of service attributes to reflect the policy being tested2) Repeat the use of the disaggregate model on all of the observations in the representative sample

Individual choices

x BEFORE = Expansion factor = Base Case ResultUse aggregation scheme to factor up results to scale of total population

New individual choices

x AFTER = Expansion factor = Policy ResponseUse same aggregation scheme to factor up results

For each project or policy to be tested:

3) Compare Base Case to Policy Response : BEFORE - AFTER = IMPACT

Alterrnative service offerings described in terms of their level of servce attributes: Waiting time Transit time Reliability Transport charges

Disaggregate data is required for the model. Though disaggregate data are difficult to find a few disaggregate data bases do exist and there are ways that certain of the aggregate data can be manipulated to serve as the input data needed to run this model.

1. Serial Number2. Commodity Description3. Commodity Code—Standard Transportation Commodity Code*4. Pounds per Year*5. Pounds per Shipment*6. Value of Commodity—Dollars per pound*7. Origination State8. Destination State9. Origin FIPS10. Destination FIPS11. Observed Mode (Truck)*12. Truck rate per mile for 3S2*13. Truck highway miles*14. Truckload per shipment*15. Number of Trailer on Flat Car (TOFC)/Number of Container on Flat Car (COFC) (0)*16. Rail Junction Frequency (0)*17. Observed Rail revenue per hundred weight (cwt) (1)*18. Rail variable cost per cwt*19. Rail miles*20. TOFC pickup miles*21. TOFC delivery miles*

Items denoted with an asterisk are required and default values are noted in parenthesis.

The disaggregate data is typically documented in a paper/electronic record of the movement such as freight bill which shows the date of the shipment, the name of the shipper, the name of the receiver, the origin and destination, the size of shipment, the mode, the freight changes, and any special handling requirements etc. What is not typically available is the level of service variables that prevailed at that time (which needs to be inferred from mileage, the conditions of transit etc). Also the annual usage of that product by the receiver is not available in the data which is an essential input required for the model.

Note that the type of disaggregate data required for the analysis depend on the policy question. If the question is about truck-to-rail diversion due to improvement in rail service then we use a representative sample of individual truck movements. If the question is about rail-to-truck diversions then we use a representative sample of rail movements.

Where can we get disaggregate data?

For rail carload and intermodal data, one of the best existing disaggregate database is the annual Rail Carload Waybill Sample compiled by the Surface Transportation Board. It contains many of the information required such as rail station of origin and destination, name of the rail carriers, route traveled, mileages, interchange points, car type used, size of the shipment, rail charges and variable costs.

For disaggregate data regarding truckload movements Freight Analysis Framework can be used which provides truckload movements between a given county of origin and a county of destination at a very detailed commodity STCC level for a particular truck body type and axle configuration.

Developing the Other InputsIn general, for the observed mode, some of the inputs, including the origin and destination, will be available from the disaggregate data; while for the alternative mode, the entire set of inputs will need to be developed. (The alternative mode would use the same origin and destination as the observed mode.) However, a number of the inputs to the ITIC model are typically not available from the disaggregate movement records. These include the highway mileage, intermodal pickup and delivery (drayage) distances, the principal commodity attributes (lbs/cu ft, $/lb), the truck payload and rates, and the rail variable cost/cwt. If truck is the observed mode and the Freight Analysis Framework is used as the disaggregate input, all of the alternative rail mode input, will need to be developed. It should be noted that the ITIC model (as it is presented here) uses mileages and a set of predefined relationships to develop many of the level of service variables, including the waiting time, the transit time and the reliability of the arrival for each of the modes based upon parameters in the model.

Required Inputs to the ITIC-IM Model (version 1.0)

# Input Definition Source Value

1 Description Commodity Description TRANSEARCH Shipment

Dependent

2 STCC Code for commodity being carried in the shipment TRANSEARCH Shipment Dependent

3 Lbs/yr Total annual weight of a given commodity transported between regions (2002)

TRANSEARCH Shipment Dependent

4 Lbs/shipment Factor for ensuring that weight per shipment is not over legal limits or cubic footage of truck trailer or COFC

TRANSEARCH Shipment Dependent

5 Dollars/lb Average value of a given commodity class TRANSEARCH Shipment Dependent

6 Ostate 2-digit U.S. state of Origin of move TRANSEARCH Shipment Dependent

7 Dstate 2-digit U.S. state of Destination of move TRANSEARCH Shipment Dependent

8 Ofips FIPS number of the county of origin of move TRANSEARCH Shipment Dependent

9 Dfips FIPS number of the county of destination of move TRANSEARCH Shipment Dependent

10 Obs mode The mode on which the shipment was observed to be traveling TRANSEARCH Truck

11 Cost/mile 3S2 Per-mile cost of using a truck for shipping ??? Shipment Dependent

12 3S2 Miles Average distance traveled by truck Use Ofips – Dfips distance??

Shipment Dependent

13 3S2 Load Maximum payload weight allowed on the truck subject to the legal restrictions

Use value of input #4

Shipment Dependent

14 num TOFC/COFC If observed mode is rail, num TOFC/COFC is “1” for rail carload, “2” for TOFC or “3” for COFC

Not Applicable Shipment DependentAll 1 in sample

15 Junction Frequency

Used to accommodate a railroad junction. If rail service may require an interchange between railroads.

Distance Dependent

Shipment Dependent0 = no junction 1 = a junction

16 Obs rail rev per cwt

Rail revenue per car divided by shipment size in hundredweights, $/cwt, (cwt =100 lbs). Rail rate is calculated as 95% of truck rate.

NA (Will be computed)

Shipment Dependent1 to start with

17 rail VC per cwt Rail variable cost per hundredweight. A simple formula is (VC per mile * Rail Miles + Lift cost at orgin at destination) * 100 / Shipment size in lb

ITIC Model Shipment DependentA simple formula is provided

18 Rail miles Distance traveled on rail FRA lookup table???

Shipment Dependent

19 Pickup miles Distance from shipper’s loading dock to origin intermodal FRA lookup table???

Shipment Dependent

20 Delivery miles Distance from destination intermodal terminal to receiver’s loading dock

FRA lookup table???

Shipment Dependent

21 Pounds per cubic foot

Weight per cubic foot of a given commodity class ITIC Model Commodity Dependent

22 Service percent Probability of no stock out (inventory) during the replenishment cycle

ITIC Default Commodity Dependent

23 Inventory carrying cost %

Costs associated with possession of a commodity, including capital cost, insurance, taxes, obsolescence, pilferage, transfer, handling, and storage.

ITIC Default Commodity Dependent

24 Order cost Cost of placing an order to be shipped ITIC Default $30

25 Interest Cost of capital during transit and during loss and damage claims ITIC Default 10%

26 Rail mph 24 hour Average rail speed ITIC Default 30 mph

27 Dray mph Average speed of drayage to/from rail ITIC Default 30 mph

28 Truck mph Average truck speed ITIC Default 50 mph

29 Load and unload hours

Amount of time needed to load or unload a truck trailer or COFC

ITIC Default TOFC = 0.5 hours, Truck = 0.5 hours

30 Hourly wage Wage paid to workers loading and unloading a truck trailer or container on a flat car (COFC)

ITIC Default TOFC = $20/hr, Truck = $20/hr

31 Pickup charges per shipment

Flat fee charged to pick up shipments by rail ITIC Default $125

32 Delivery charges per shipment

Flat fee charged to deliver shipments by rail ITIC Default $125

33 Pickup charge per mile

Per-mile charge for rail drayage over 30 miles at origin ITIC Default $1.38 for every mile above 30 miles

34 Delivery charge per mile

Per-mile charge for rail drayage over 30 miles at destination ITIC Default $1.38 for every mile above 30 miles

35 Dwell time at Terminal

Number of days the shipment stays at the origin and destination intermodal terminal

ITIC Default 0.5 days at origin and 0.5 days at destination

36 Interchange delays If rail service may require an interchange between railroads ITIC Default 1.5 days

37 Wait time Number of days before a shipment can be transported ITIC Default TOFC = 0.5 days, 3S2 = 0.5 days

38 Reliability A factor used to represent the variance of the transit time for truck and rail

ITIC Default TOFC = 0.45, 3S2 = 0.40

39 Loss and damage as a percentage of gross revenue

Ratio of loss and damage costs to commodities over the gross revenue from shipping

ITIC Default TOFC = 0.20%, 3S2 = 0.07%

40 Claim Payment Days

Days required to process a loss and damage claim ITIC Default TOFC = 90 days, 3S2 = 30 days

41 Additional Fee per Mile

If there is an additional cost to for truck per mile in addition to ‘Cost/mile 3S2’

ITIC Default $0 / mile

42 One time additional fee

If a fixed one-time fee applies for truck (special license or single use fee)

ITIC Default $0

43 Rail Rate per cwt 0.95 * (Cost/Mile 3S2 * 3S2 Miles) * 100 / Shipment size in lb ITIC Default 0.95 * Truck Cost

44 Dunnage Extra charge to rail orders ITIC Default $50

Entering the data in ITIC Model

Once the disaggregate data is available it is entered in the sheet1 of truck itic input.xls which is the data file for the model. The disaggregate data must be entered in the similar format as the sample entries. Each row indicate one record/shipment and there are 21 parameters for each record. Some inputs are descriptive and are not essential for determining logistics costs. Such inputs can be left out or some value or text can be used to occupy the cell. However, to maintain the detail of a move, no data should be excluded.

Running the model

There are three buttons in the first worksheet named “Running Macros” in the model which are used to perform specific tasks within the model using macros. A chosen macro reads a record from “Sheet1” of data file and writes it into the “TSW” worksheet of the ITIC model. It then writes the results to another spreadsheet in the data file for further analysis.

A comparison of logistics costs per year are the determining factor for which mode wins the business. However, there are caveats or conditions that must be met.

A shipment is considered a candidate for diversion to rail if both of the following is true-

i) If Rail Variable Cost is below Rail Revenue. (CellD85) In other words, 110 percent of rail variable cost plus dray cost must be less than the calculated rail rate. The logic is a railroad will not carry the traffic at a rate below cost.

ii) If Rail saves more than 3% in transportation and logistics cost compared to Truck (Cell D83) OR There is a cost savings of $20,000 per year by Rail compared to truck (Cell D84). The logic is that for truck to divert to rail, a shipper/receiver had to reach a hurdle in saved logistics costs. This approach was taken to avoid instances where diversion would occur if rail logistics were 1 cent below truck logistics costs but still offering a realistic hurdle on savings. )

Contents of Truck itic input.xls

“Sheet1” – It contains truck flow data for input. “Sheet2” – It will contain clean data void of mis-assignments following the “Base Case” run. “Sheet3” – It will contain the logistics cost and other statistics generated by the model for the records in the base case. The user simply clicks the “Base Case” button in the model.

“Sheet4” - It will contain the diversion results for review. Following service parameter changes, the user simply clicks “Policy Case” button in the model.

Contents of the model ITIC_IM v1.0.XLS

“Running Macros” Sheet- It provides the user to perform base case and policy scenarios analysis by simply clicking buttons. This sheet also provides summary statistics, which allow side-by-side comparison of the Base as well as the Policy Case for both truck and rail intermodal.“TSW” Sheet- This sheet is the heart of the model. Variables from the disaggregate data file and the input parameters are used to compute the level of service components of the shipper’s logistics cost function for the observed mode and for each of the alternatives. These, in turn, are used to determine the selected mode.“Assumptions” Sheet- This sheet contains not only many of the assumptions affecting level of service variables for the base and policy runs, but two important tables—the Service Level Parameter Table (Table 1) and the table of Mode-Equipment Options and User Supplied Input Parameters (Table 13).“Rail” Sheet- In this sheet the various elements involved in determining the freight charges for rail are assembled to determine the observed charge per shipment Rail freight charges are computed in a table on this spreadsheet. A side calculation of the total variable cost including dray charges is included in this sheet. Also it contains a simple rail variable cost calculator, if the user is unable to obtain reliable rail line haul variable cost.“Truck” Sheet- Truck freight charges are computed in a table on this spreadsheet.“Assumptions Default” Sheet - This worksheet contains all the default values that were provided for in the original ITIC-IM model. FRA and FHWA provide default values as guidelines; users are strongly urged to obtain data relevant to their analysis issue. Formulas in TSW worksheet

Transportation Cost

Lb/day = (lb/year) / 365

Days between orders = final lbs in the shipment / (lb/day)

Transit time (Rail) = Rail Miles / Rail Speed + Dray Miles / Dray Speed + Dwell time at origin and destination terminals (0.5 days at each) + Interchange delays (1.5 days – if there is an interchange) [Both rail and dray speed is based on 24 hours]

Transit time (Truck) = Truck Miles / Truck Speed [truck speed is based on 10 hours a day]

Expected L&D claim per shipment = L&D as percent of gross freight revenue X transport charges per shipment

Transport charges per shipment (Rail) = (Load/unload hours X Hourly wage) + final lb in shipment X rail rev per cwt / 100

Transport charges per shipment (Truck) = Handling cost per shipment + Linehaul cost per shipment = (Load/unload hours X Hourly wage) + (Linehaul miles X Linehaul cost per mile + Linehaul miles X Additional fee per mile + One time additional fee)

Number of shipments/yr = final lbs in shipment X lb/year

transport charges/yr = transport charges per ship X Number of shipments/yr

Non Transportation Logistics Costs

Order cost = (Order cost per shipment + Dunnage) X Number of shipments/yr [for ‘truck’ dunnage is 0]

In-transit stock carrying cost = Lb/day X Dollars/lb X Transit time X Interest

cycle stock carrying cost = Dollars/lb X inventory carrying cost factor X final lbs in shipment / 2

Loss & damage claims = Expected L&D claim per shipment X Number of shipments/yr

Capital cost on claims = claim payment days X loss & damage claims X Interest

Safety stock carrying cost = safety stock (lb) X Dollars/lb X inventory carrying cost factor

Total Non-Transportation Logistics Costs = Order cost + In-transit stock carrying cost + cycle stock carrying cost + Loss & damage claims + Capital cost on claims + safety stock carrying cost

Rail Revenue per cwt

0.95 X Cost per mile of 3S2 X 3S2 Miles X 100 / Shipment weight in lb

Rail Variable Cost

Dray Pickup charge / Shipment weight in lb X 100 + ((Variable Cost per mile X Rail Miles + Total Lift charges at Origin and Destination) / Shipment weight in lb X 100) / 0.909

Case Study

We use SC Transearch 2008 as preliminary data source for highway freight movements. We select the shipments forcasted for year 2023 to be originating at South Carolina and destined to state of Florida and Virginia. There are about 22000 full truckload entries in the database with STCC Commodity codes at 4 digit level.

To perform the analysis we need a forecast of 2023 freight rate. An analysis of the operational costs of trucking is published in 2012 by American Transportation Research Institute (ATRI). The following two tables are from their report (http://www.glostone.com/wp-ontent/uploads/2012/09/ATRI-Operational-Costs-of-Trucking-2012.pdf). Only four previous year’s data are available.

Inadequate historical data and long forecast horizon makes forecasting difficult and uncertain. Extrapolation techniques have higher success in short time horizon. Regression, time series analysis options can be pursued. In the sample case study a value of $2.10 per mile is assumed for now.

Also, SC 2008 Transearch does not provide county outside SC, only Bureau of Economic Areas (BEAs). The Center for Transportation Analysis (CTA) in the Oak Ridge National Laboratory developed data that contain a matrix of distances between each pair of county centroids via highway, railroad, water, and combined highway-rail paths (http://cta.ornl.gov/transnet/). A possible option is to obtain as estimate of intermodal railway miles based on highway miles. The result of regression between intermodal rail and highway miles based on 125 samples is provided below.

0 500 1000 1500 2000 2500 3000 3500 40000

500

1000

1500

2000

2500

3000

3500

4000

f(x) = 1.04308599789475 x + 67.4614019211329R² = 0.976919633181141

Rail Miles Vs Highway Miles

To estimate dray miles we can use the same CTA database and take dray miles as a percentage of rail miles and then use data fitting. The best fit distribution is ‘Gamma’ out of seven distributions (beta, normal, uniform, logistic, gamma, weibull, lognormal) tested using ‘fitdistrplus’ package in R (Anderson-Darling statistic: 0.3201 ) based on 125 samples. Fitting of the distribution ' gamma ' by maximum likelihood

Parameters :

estimate Std. Error

shape 1.3586393 0.15588283

rate 0.1922481 0.02656798

We can use GAMMA.INV(RAND(),1.359,1/0.192) * Rail Miles / 100 in EXCEL to estimate of Dray miles based on intermodal rail miles.

The next step is to populate the input file truck itic input.xls. We need to perform some data cleanup such as discard Transearch records when ‘shipment weight’ is zero and when ‘truck miles’ is zero. Also, we ensure to format ‘STCC’ codes as ‘Text’ in Excel, not as ‘Numbers’ as desired in the model. A zero before one digit STCC code to make it two digits (e.g. 01, 08, 09 instead of 1, 8, 9). We put ‘unknown’ in ‘rail VC per cwt’ column and ‘1’ in ‘obs rail rev per cwt’. For ‘Junction Frequency = 1 if Truck Miles > 1000, otherwise 0.

About 17000 records were selected for policy analysis after base case run of about 22000 records. There were not many Mis-assigned, mostly excluded because they failed ‘cost < revenue’ criteria. We assume by Year 2023 Rail is offering the following service improvements-

• Increase rail speed by 8 mph (30 mph to 38 mph, Truck = 50 mph)

• Improve reliability from 0.45 to 0.42 (Truck = 0.40)

• Reduce Loss & Damage Claims 20% to 10% (Truck = 7%)

Results and Discussion

The output of policy analysis is shown below.

TRUCK-TO-RAIL INTERMODAL OUTPUT SUMMARY STATISTICS

Base Case Policy Case Results

(Truck) TruckDiversion

Number of Records 17,733 17,733 0

Number of Shipments 314,473 314,473 0

Tons Shipped 6,004,312 6,004,312 0

Truck VMT148,171,808

148,171,808 0

Intermodal Dray VMT 0 0 0

Rail Ton-Miles 0 0 0

Logistics Costs($) 451616829 451616829 0

It shows that given above rail service improvements may not be enough for some truck shipments to be converted to rail. The benefits of rail service improvements is more likely to affect longer transportation distances, however. The input and output for a shipment from Spartanburg, SC to Fort Meyers, FL is shown below. The shipment was not converted to rail because the 3% or $20000 threshold was not exceeded.

Conclusion

This class project applies the ITIC-IM model to perform modal diversion and economic benefit analysis. Formulas underlying the model to calculate logistic costs are presented. For our truck to rail diversion we used Transearch Database as our primary data sources. Whenever data was not available some methods were proposed that can be used to estimate the input. A case study was performed on freight movements originating at South Carolina and destined to state of Florida and Virginia.

Serial number 20036Description In bulk in boxcarsSTCC 50lbs/yr 48689540lbs/shpmt 41116.49784$/lb 0.563500086Ostate 45Dstate 12Ofips 45083Dfips 31Obs mode truckCost/mile 3S2 2.13S2 miles 714.23S2 load 41116.49784num TOFC/COFC 1Junction Frequency 0obs rail rev per cwt 1rail VC per cwt unknownrail miles 749.91tofc pu 18.74775tofc dlvr 18.74775

InputSerial Number 20036STCC 50Ostate 45Dstate 12pounds/yr 48689540rail miles 749.913-S2 miles 714.2chosen mode 2line haul miles per shipment final 714.2annual pickup cost 0annual delivery cost 0 rail ton-miles 0translog per cwt chosen mode final 3.830437transport cost per cwt chosen mode final 3.672054payload final 41116.5number of shipments 1185Intermodal Dray VMT 03S2 translog cost per cwt 3.8304373S2 transport cost per cwt 3.6720543S2 annual VMT 8463273S2 number of shipments 1185annual rail transport cost 1700272annual rail non transport cost 147410.1origin fips 45083destination fips 31Calculated Rail Rate/cwt 3.465346New Rail Revenue 0annual truck transport cost 1789137annual truck non transport cost 77115.82total annual rail logistics cost 1847682total annual 3S2 truck logistics cost 1866253flag for rail rate less than costs SAVINGS 18570.09

OUTPUT