Introduction to TrackShare

ZETA-TECH Associates, Inc. is a US consulting firm specializing in the railroad and rail rapid transit industries. In 1987, ZETA-TECH (ZT) was asked by a client within the North American railroad industry to develop a costing methodology for four tasks:

q Properly allocate maintenance-of-way and renewal costs between heavy axle load traffic (e.g. unit trains), moderate axle load, high speed traffic such as intermodal trains, and high speed passenger trains;

q Reflect more accurately than historical methods the cost/density relationship in maintenance-of-way expenses;

q Be “auditable” by regulatory bodies;

q Take account of modern research into track costs and the effect of axle loads, speeds, and other traffic variables.

The original model was named the Weighted System Average Cost Model (WSAC). After a large number of enancements and changes the model is now reffered to as TrackShare. TrackShare relies on engineering equations to “weight” the ton-miles of each type of train traffic in proportion to the damage each type causes to the right-of-way assets.

TrackShare uses “Engineering Adjustment Factors” (EAFs) to quantify the relationship between traffic volume and track maintenance cost for any desired number of defined traffic types, based on vehicle characteristics such as axle load, speed, and bogie type, and track characteristics such as grade, curvature, and weight of rail.

This is the first model to use engineering equations to determine costs in a manner defensible before a regulatory body. The tool won ICC approval in a 1995 passenger case involving Amtrak and Conrail.

TRACKSHARE has been used by six large North American railroads. The most common application for TRACKSHARE has been the development of incremental MOW costs associated with passenger train operation. TRACKSHARE has also been incorporated into the managerial costing systems of two North American railroads, and has been used by four others. One railroad distinguishes between ten types of freight and passenger traffic.

TRACKSHARE is grounded in engineering research. The Engineering Adjustment Fs are based on known engineering relationships, as determined by tests and years of research. TRACKSHARE does not itself generate an expenditure number. Instead, the TRACKSHARE model calculates “relative” damage to the track structure for each traffic/track segment combination, and uses this relative damage to adjust against either system average maintenance expenditures or precise segment expenditures. System average data is useful when geographic specific records have either not been recorded by a track owner or when the track records are not long enough to establish a normalized allocation.

TRACKSHARE is applied only to those components directly affected by the passage of trains (rail, ties, and ballast). These variable expenses typically comprise about one half of total permanent way maintenance costs. The remaining 50% of track maintenance costs do not vary directly with traffic.

“Economies of Density”: Are There Limits?

A significant issue surrounding rail costing and setting of track occupancy and use prices is the question of Economies of Density. Old formulas and management theory said that “the incremental costs” of maintenance continued to decline as traffic volume or density increased. A model for representing that approach in regulatory proceedings and rail-to-rail negotiations – called the Speed Factor Gross Tons formula – was relied upon to generate allocated costs. ZETA-TECH challenged that assumption.

ZT demonstrated that when total train volume increases beyond 25 million gross tons, track maintenance costs increase in a linear relationship with traffic increases. This means that the traditional railroad and ‘regulator’ treatment of MOW costs as fixed or declining with tonnage increases has been wrong. TRACKSHARE measures the real impact of direct damage related to different train categories on heavily used rights-of-way.

ZT proved that as the total density of any segment declines below 25 million gross tons of traffic per year, the deterioration of track components is strongly affected by environmental considerations. Therefore, at extremely low track densities, segments tend to be dominated by costs of structure repair, signal maintenance, snow clearance, and vegetation control. The pattern on light density lines is dominated by these fixed or programmed expenditures rather than tonnage and speed related vehicle damage.

The section below describes an approach to employing the TRACKSHARE tool.

How TrackShare is Used

Around the world, shippers, third party intermodal operators, passenger train operators, freight carriers with trackage rights on another carrier, and track owning companies are all concerned with questions of equity for users and owners. The terminology is different from place to place: “open access”, trackage rights, and “competitive access”. The need for unbiased measurements is universal. Given a situation where multiple users are in dispute over the share each must bear for joint costs, how would ZETA-TECH’s TRACKSHARE model be useful? How would the tool be deployed?

Step by Step Methodology

The first step is to develop appropriate local condition "engineering adjustment factors" (EAFs) for the “joint use” track segments with ZETA-TECH equations. These equations have been validated by recent research test data, and have been reviewed by independent engineering organizations. ZETA-TECH will calibrate its equations to the local “joint use” conditions. Local track data and train/vehicle category data must be collected for the specific geographic segments.

Track Geography Data

Joint-use tracks must be divided into a network of unique track segments, without overlapping. For each of these segments, data must be obtained on the following parameters:

q Grade (in percent)

q Curvature (in degrees or by radius)

q Weight of rail (per yard)

q Rail type (welded or bolted)

q Rail hardness

q Level of rail lubrication (none, moderate, heavy)

q Rail fastening system

q Sleeper type (concrete, steel, wood)

q Track miles per route mile.

q Traffic density per track mile for each defined traffic type

q Total track miles, identified by type including controlled sidings

q Turnouts per mile

Track segments may be of any length that can be supported by track data records. A multi-year time series of permanent way operating expense and capital additions is recommended for use in the analysis. In U.S. applications of TRACKSHARE, costs from U.S. Interstate Commerce Commission R-1 reports, Schedules 330 and 410 have been successfully used. Specific costs developed from internal, non-published accounting data can also be used.

Train and Vehicle Data

For each defined traffic type, each geographic segment must obtain the following data.

q Gross and tare weight

q Axle load

q Number of axles per wagon

q Type of rail-truck or bogie (self-steering, three-piece, etc.)

q Wheel diameter

q Wheel profile

q Wheel tread condition

q Average operating speed (from timetable or from simulations)

q Train length (number of axles)

Rail vehicle data is generally available from plans, bills of lading, and similar sources.

Any number of traffic types may be defined -- a typical number might be five types of freight traffic and three passenger categories that includes commuter rail, intercity, and country or light-rail services. For each segment, a weighted average operating speed is calculated for each traffic type based on speed limits, and any special restrictions applying to particular traffic types. For example, unit trains of coal might be restricted to a maximum of 35 mph while other trains move at higher speeds.

Curvature Calculations

To calculate average curvature for each segment, the following process is employed. First, track-miles are accumulated in appropriate curvature categories. Since the effect of curvature is non-linear, sufficient categories must be defined to produce an acceptable linear approximation when curvatures are averaged. The following curvature categories are typically used:

0 - 1°

> 1° to £ 3°

> 3° to £ 5°

> 5°

Track with curvature of 1° or less is treated as tangent track. Curvature in the second category is assumed to have a constant curvature of 1°, track in the third category a curvature of 3°, and track in the final category a curvature of 5°. Since curvature has a non-linear effect in the damage equations, this grouping and averaging is a reasonable compromise between accuracy and ease of computation, as long as a sufficient number of categories are defined to capture the full variation in curvature.

Rail weight and the absolute value of grade (expressed in percent) are averaged over each track segment, since their effects are linear in the track damage equations. Lubrication is taken as a single value for each segment. Rail hardness (if different within a segment) is linearly averaged.

Percentages of continuous welded rail (CWR) and bolted rail are calculated for each segment. These percentages are multiplied by ZT factors reflecting the relative damage effects on welded vs. bolted rail to calculate the appropriate segment values.

EAF Calculations

To properly account for differences in maintenance costs between track segments, when segment-specific cost data is unavailable, the TRACKSHARE model calculates weighted system average EAFs for each of the defined traffic types, using system average values for track characteristics. Then these system average EAFs for each traffic type are adjusted to reflect differences in track characteristics among segments. The result is to produce higher EAFs (and therefore higher costs) on some track segments, and lower EAFs and costs on others, depending upon grade, curvature, and weight and type of rail. Thus, the cost of each traffic type will be different on each segment, and average cost on each segment will not necessarily be equal to the system average cost.

An EAF less than one (1) indicates that traffic on that segment causes damage less than average, while an EAF greater than one indicates more than average damage.

Output Example

The following section describes how TRACKSHARE takes different geographic characteristics and calculates appropriate costs to the various railway users. The numbers are based on an actual case study.

Table 1

CASE STUDY GEOGRAPHY

Geographic Link Distance Track Characteristics

Segment 1 10 miles Heavy curves and grades, all heavy & welded rail

Segment 2 20 miles Tangent track, level grade, all heavy & welded rail

Segment 3 30 miles Tangent track, level grade, all light rail, and bolted

The ZETA-TECH Damage Factors are calculated for each segment. Note that a railway’s system wide basis, average traffic of all types operating on all different track characteristics, would have a damage factor of 1.0. Using the data for Segment 2 (above) ZETA-TECH produces the following damage factors for four very different train or business types.

Table 2

DAMAGE FACTORS, BY TRAFFIC TYPE

Business User MGT/Yr Link 1 Link 2 Link 3

Loaded Unit Trains 18 1.77 1.31 1.53

Empty Unit Trains 6 .80 0.69 .75

Passenger Trains 1 1.32 1.03 1.17

Mixed Freights 10 1.52 1.16 1.33

- Each business or traffic type has its own speed, equipment type and other parameters (rounded)

Table 3 identifies the calculation of weighted and unweighted densities for the three links as GTM. The total ton miles is 2,100 and the weighted ton miles is 2,734 in this case study. A system average cost of $0.0012/GTM is obtained from the track owner’s records.

For all three track segments in this case study, total maintenance costs from company records were $2.5 million. How should this cost be allocated to the different users? First, the damage factors must be transformed into Engineering Adjustment Factors (EAF’s), a process that ensures the total of adjusted costs is equal to the total of actual costs. Table 4 identifies the use of EAF’s, by business user type.

Table 3

GROSS TON MILES BY SEGMENT

Business User Link 1 Link 2 Link 3 Total GTM

Loaded Unit Trains

Unweighted 180 360 540 1,080

Weighted 319 472 826 1,616

Empty Unit Trains

Unweighted 60 120 180 360

Weighted 48 83 135 257

Passenger Trains

Unweighted 10 20 30 60

Weighted 13 21 35 69

Mixed Freights

Unweighted 100 200 300 600

Weighted 152 232 399 783

- Table numbers have been rounded

Table 4

Engineering Adjustment Factors

Traffic Type EAF M-O-W Unit Cost Total M-O-W Cost

Loaded Unit Train 1.150 $ 0.0014 $1,489,800.

Empty Unit Train 0.567 $ 0.0007 $ 245,000.

Passenger Train 0.882 $ 0.0011 $ 63,500.

Mixed Freight 1.002 $ 0.0012 $ 721,700.

TOTAL $2,520,000.

Sample calculation for Passenger train case is: (2100*69)/(2734*60) = .882

Sample calculation for Loaded unit train is: (2100*1616)/(2734*1080) = 1.15

Table 5 identifies the cost allocation for this sample sixty miles of track, on a percentage basis, after the completed ZETA-TECH analysis.

Table 5

Percentage Cost Share for Each Business User

Traffic Type Tonnage System Avg. Costs TRACKSHARE

Loaded Unit Train 18 MGT 51.4% 59.1%

Empty Unit Train 6 MGT 17.1 9.7

Passenger Train 1 MGT 2.9 2.5

Mixed freights 10 MGT 28.6 28.6

- Intermediate calculations and methodology are omitted from this sample presentation.

With 51% of the line tonnage moving across all three of the case study track segments, the heavy unit trains are allocated 59% of the maintenance costs. Passenger trains are allocated 2.5% of the costs at 2.8% of tonnage.

These conditions and allocations were calculated using actual data. The site and the specific details remain confidential, as are the actual ZETA-TECH proprietary formulas. The tables nevertheless illustrate sensitivity to real localized inputs and a diverse set of passing traffic and equipment. Actual prices to be charged relative to these basic cost allocations would in turn vary by market conditions, slot-times for occupying track capacity, and other factors such as embedded asset value of the permanent right-of-way.

Conclusion

ZETA-TECH’s TRACKSHARE model permits a track owner and both current users and potential new users to base fees upon actual conditions of wear and tear. It is thus possible to establish an equitable solution for “open access” solutions and traditional Trackage Rights Agreements, based on the geographic specifics of incremental joint use. Of course, additional calculations with recognized economic formulas must be made to account for return on assets, and other related expense. A separate module is used, for example, to calculate adjustments for “solely attributable” costs.

For any given business case, TRACKSHARE model output is based on engineered values covering:

q Increases in rail wear and rail fatigue damage

q Gage Widening and tie wear

q Degradation of track geometry

q Differences caused by traffic type or Business User

TRACKSHARE will run on an IBM compatible personal computer with 100 Mhz clock speed.

As public discussion increases about rail mergers, service failures and related “open access” issues, the TRACKSHARE model can provide economic numbers to measure proposed solutions. Track owning companies like Railtrack in the UK also have a long-term financial stake in these issues.

Finance Docket No. 32467, For An Order Fixing Just Compensation, Dec. 29, 1995