Table of Contents

• Introduction
• Canadian National
• Canadian Pacific
• BNSF
• CSXT
• Norfolk Southern
• Union Pacific

Introduction

These methodologies have been prepared by the various railways and reflect how the various performance indicators have been calculated. Neither Transport Canada nor Statistics Canada are responsible for the contents of this document. The following methodologies were translated from English into French by Transport Canada and reviewed by the respective railways.

Canadian National

1. Weekly Average Train Speed by Train Type
   • Industry train speed used (as submitted to the AAR), expressed in Kilometers per hour.
   • Data is provided by AAR week (Saturday 00:01 to Friday 23:59).
   • Calculated by summing the total train kilometres, and dividing by the total train hours operated.
   • Includes system trains running on Company lines, and system trains operating on non-system lines under trackage/running rights or as a detour (e.g. system trains operating on CP lines in the Fraser Canyon).
   • Excludes passenger, work and company service trains, yards, transfers, local trains, and road switchers. Excludes non-system trains operating on company lines under trackage/running rights or as a detour. Excludes crew change time, and other terminal time at crew change points.
   • Reported by train type (Manifest 200-400 series, I/M 100 series, Grain G800 series, Coal C700, Crude Oil (specific train ids) and Ethanol (specific train ids)).
   • Cross border trains are included immediately upon crossing the border. Mileage between reporting stations, i.e. Line Segments, are split to country and province and is aggregated to calculate total train miles, and total train hours operated.
2. Weekly Average Dwell for our 10 largest Canadian terminals
   • Industry Process Dwell used (same as used for AAR performance report), expressed in hours per car.
   • Data is provided by AAR week (Saturday 00:01 to Friday 23:59).
   • The ten terminals selected for the report are those which have the largest number of cars processed. This same list of stations will be reported each week.
   • Average elapsed time from entry to exit of cars processed at major terminals (excludes cars on • through trains).
   • Entry events include train arrival, interchange receipt and customer release.
   • Exit events include train departure, interchange delivery and first placement.
   • Excludes cars placed into hold, storage or bad order, or company service cars.
   • An articulated car is counted as one regardless of the number of platforms.
3. Weekly Average Cars On-line by Car Type
   • Weekly average of active cars online per day, calculated by averaging the daily car inventory (04:00 ET).
   • Data is provided by AAR week (Saturday 00:01 to Friday 23:59).
   • Active cars include cars on trains, in yards and at customer.
   • Includes CN owned and leased, private equipment and foreign equipment on-line.
   • Excludes stored cars including private cars in chargeable storage.
   • Excludes heavy bad order, dismantlers, company service, and passenger cars.
   • Excludes Private Cars placed on customer siding for more than 30 days.
   • An articulated car is counted as one regardless of the number of platforms.
4. Weekly Average Dwell at Origin for Unit Train Shipments (grain, coal, auto, crude, ethanol and other unit trains)
   • Average time for cars in normal status measured from customer release to train departure.
   • Data is provided by AAR week (Saturday 00:01 to Friday 23:59). Excludes time for cars placed in chargeable hold.
   • Limited to B, C, G, S, U trains with more than 70 cars with same origin and destination.
   • (B-Potash, C-Coal, G-Grain, S-Sulphur and U-Other Unit trains – Crude/Ethanol).
   • Crude trains identified by specific train ids.
   • Release and train departure station must be the same.
   • Excludes time placed into chargeable hold (EX) between release and departure events.
5. Weekly Average Trains Held Short of Destination
   • Weekly average count of late trains per day that are held at snapshot time (00:01 ET).
   • Data is provided by AAR week (Saturday 00:01 to Friday 23:59).
   • Held trains are defined as exceeding the scheduled time at the snapshot location by 1 hours.
   • Delays are broken down by cause (Crew, Locomotive Power, or Other).
   • Reported by train type (Manifest 200-400 series, I/M 100 series, Grain G800 series, Coal C700,
   • Crude Oil (specific train ids) and Ethanol (specific train ids)).
   • Delays caused by other reasons could include: disruption, weather, foreign, engineering, equipment, or passenger delays.
6. Weekly Average Cars Unmoved for Over 48 hours
   • Weekly average count of cars per day in normal status in a yard, or tendered to CN on a customer track, that have not moved in over 48 hours.
   • Data is provided by AAR week (Saturday 00:01 to Friday 23:59).
   • Excludes cars that are placed at a customer, stored, in heavy bad order status, dismantlers and OCS cars.
   • Commodity is calculated by waybill Major Sub Group (a subset of Business Unit), except
   • Intermodal which is by Car Group, and Ethanol by STCC.
   • Uses daily snapshot at 23:50 ET.
7. Weekly Total Grain Cars Loaded and Billed, Aggregated and by Province
   • Cars measured on waybill activation with Grain commodities (Business Unit Grain, see appendix for full STCC list).
   • Data is provided per Grain week, from Sunday 00:01 to Saturday 23:59
   • Province provided is the waybill CN origin.
   • Includes System-controlled and Private-controlled cars, and including all car types except Intermodal.
8. Running Total Number and Average Days Late for Outstanding Grain Car Orders; Total New Car Orders, Filled and Cancelled During Past Week
   • Includes System-controlled covered hopper cars only.
   • Western Canada is managed by the Bulk Operations group.
   • Eastern Canada is managed by CCO as manifest orders.
   • Data is provided per Grain week, from Sunday 00:01 to Saturday 23:59
   • Western Canada orders reported are sourced from the Planned Service Report (PSR), set on the
   • Thursday prior to the order week, which excludes early customer cancelations and orders without terminal authorization. Eastern Canada orders are net orders, orders less cancellations.
   • Filled orders include add in’s, and cars spotted for the following or prior weeks orders.
   • An order is considered to be overdue if a car has not been supplied by Saturday 23:59.

Canadian Pacific

BNSF

Methodological Report of BNSF Railway Company

Pursuant to s. 77(4) of the Transportation Modernization Act, S.C. 2018, c.10 (the "Act"), BNSF Railway Company ("BNSF") provides the following explanation of the methodology used to generate the periodic reporting required under ss. 77(2)-(3) of the Act. Pursuant s. 77(2), Class I rail carriers are required to provide to the Minister of Transport information on specified service and performance indicators for that carrier's network in Canada for each period of seven days. While the Governor in Council is authorized pursuant to paragraph 50(1.01)(b) of the Canadian Transportation Act to make regulations requiring Class I rail carriers to provide information for the purposes of communicating service and performance indicators to the public, ss. 77(1)-(2) of the Act provides that until such regulations come into force, Class I rail carriers are to submit a report containing the information specified in 49 C.F.R. § 1250.2(a)(l)-(8), as adapted by s. 77(3) of the Act. Further, Transport Canada has provided instruction that reports provided pursuant to s. 77 of the Act must be limited to traffic the reporting Class I rail carrier moves through Canada. Consistent with the Act, BNSF's report reflects the requirements and guidance promulgated by the Surface Transportation Board of the United States (the "STB") relating to 49 C.F.R. § 1250.2(a), including Orders in STB Ex Parte No. 724, United States Rail Service Issues-Performance Data Reporting. Covered parties are required to submit their first report under the Act on December 5, 2018. The Act further provides that each Class I rail carrier shall, in its first report, provide an explanation of the methodology it used to derive the data contained therein, including the definition of unit train used for reporting purposes. Accordingly, BNSF provides the following information about the methodology employed to generate the data included with BNSF's initial performance data report that is being simultaneously electronically submitted to Transport Canada:

• Data Element No. 1-Train Speed: Weekly system-average train speed by train type (intermodal, grain unit, coal unit, automotive unit, crude oil unit, ethanol unit, manifest, all system)

Data Element No. 1 reflects data that is currently provided by BNSF and other rail carriers through the AAR's public Weekly Performance Reports, available on the AAR website. Average speed is calculated by dividing train-kilometres by total hours operated, excluding yard and local trains, passenger trains, maintenance of way trains, and terminal time. BNSF's report (i) includes the two additional categories of "Crude oil unit" and "Ethanol unit" required by 49 C.F.R. § 1250.2(a)(l) that are not separately isolated in the AAR Weekly Performance Report, and (ii) is limited to our operations in Canada.

• Data Element No. 2-Terminal Dwell Time: Weekly average terminal dwell time for the system and ten largest terminals by railcar capacity.

Data Element No. 2 also reflects data that is currently provided through the AAR's public Weekly Performance Reports. Consistent with the Act, the AAR report and BNSF's reporting obligations to the STB, terminal dwell has been measured as the average time a car resides at the specified terminal location expressed in hours, beginning with a customer release, received interchange, or train arrival event and ending with customer placement (actual or constructive), delivered or offered in interchange, or train departure event. Cars that move through a terminal on a run-through train are excluded, as are stored, bad ordered, and maintenance of way cars. Please note that the report lists only two terminals, Vancouver, BC and Winnipeg, MB, which are BNSF's only terminals in Canada. Further, Winnipeg experiences a limited volume of the car events that mark the measurement points for terminal dwell and will have no reportable events during many seven-day periods. For periods with reportable data, the reported dwell time may vary widely between periods due to the small number of events in each period. For this reason, BNSF believes that dwell time reported for Winnipeg is not likely to be an accurate indicator of operating performance at that location and cautions Transport Canada to consider this factor when reviewing the data provided in this Data Element.

• Data Element No. 3-Cars Online: Weekly total cars on line by car type (box, covered hopper, gondola, intermodal, multilevel (automotive), open hopper, tank, other, total)

Data Element No. 3 also reflects data that is currently provided through the AAR's public Weekly Performance Reports. BNSF's report reflects the average of the daily on-line inventory of freight cars on our Canadian network. Articulated cars are counted as a single unit and cars on private tracks (e.g., at a customer's facility) are counted on the last Class I rail carrier on which they were located. Maintenance of way cars are also excluded.

• Data Element No. 4-Dwell Time at Origin for Unit Trains: Weekly average dwell time at origin for loaded unit trains by type (grain, coal, automotive, crude oil, ethanol and all other unit trains)

BNSF has populated our report with data extracted from existing internal reports identifying the time between release of a loaded unit train by a customer at origin and the departure of the train from the facility for our traffic in Canada. This approach is consistent with the definition of dwell time provided in the STB's requirements on which the Act is based. The data is sorted by the individual unit train categories identified in 49 C.F.R. § 1250.2(a)(4), as referenced by the Act; "All Other Unit Trains" includes remaining categories of unit train shipments, including rock, sand and taconite unit trains. BNSF has also isolated non-origin interchange dwell and empty units from the periodic reporting.

• Data Element No. 5-Trains Holding: Weekly average trains holding per day by train type (intermodal, grain unit, coal unit, automotive unit, crude oil unit, ethanol unit, other unit, and manifest) and by cause (crew, locomotive power and other)

BNSF has populated Data Element No. 5 with data extracted from a different internal source using the snapshot approach required by 49 C.F.R. § 1250.2(a)(5), as referenced by the Act (e.g., running a daily same-time snapshot of trains held in Canada and calculating the daily average for the seven-day period by adding up the number of trains holding from each daily snapshot and then dividing that number by seven). As BNSF has explained in the record relating to the STB requirements to which the Act refers, this Data Element captures trains held at a point on BNSF's Canadian network for numerous reasons entirely separate from railroad performance, including trains that are held as part of their routine operating plan, informed by the needs of shippers, receivers and/or connecting carriers. In addition, BNSF causation flags of "crew," "locomotive power" and "other" will continue to be applied manually by dispatchers and other operating personnel based on information available to them. Delay on a single train can be the result of several causes, but the dispatcher or operator may not be fully aware of all contributing causes and, in any event, manually selects only a single cause code, which becomes the only cause subsequently reflected in this Data Element.

• Data Element No. 6-Cars Held: Weekly average loaded and empty cars in normal movement and billed to an origin and destination that have not moved in more than 48 hours, by type (intermodal, grain, coal, crude oil, automotive, ethanol, fertilizer, and all other)

BNSF has populated Data Element No. 6 using the snapshot approach required in 49 C.F.R. § 1250.2(a)(6), as referenced by the Act, by running a daily same-time snapshot of all loaded and empty cars being used in commercial service in Canada (excluding cars that have been placed in storage, constructively placed or bad ordered, and cars being used in railroad service such as ballast and other maintenance of way trains) and calculating the daily average for the week. In populating the fertilizer metric, BNSF has used the following fifteen fertilizer Standard Transportation Commodity Codes (STCCs): 2871236, 2871235, 2871238, 2819454, 2812534, 2818426, 2819815, 2818170, 2871315, 2818142, 2818146, 2871244, 2819173, 2871313, and 2871451. Cars have been counted by reference to the underlying rail equipment without accounting for how many individual units may be carried on a single piece of rail equipment. For example, an intermodal railcar will count as a single car even though it may carry multiple units (e.g., containers) at various points along the route. As with Data Element No. 5, BNSF has calculated the daily average for the week by adding up the number of qualifying cars holding from each daily snapshot and then divided that number by seven. It should also be noted that just because a car has been held at a point on the BNSF network for more than 48 hours does not mean that the car will not be delivered in a timely manner or even within the initial service plan - many cars are held in terminals and other locations on our network as part of the service design for the movement or for the convenience of a shipper or receiver. As with Data Element No. 5, potentially significant numbers of delays that are not linked to BNSF's own service performance will be captured as BNSF delays in the data reported pursuant to this Data Element.

• Data Element No. 7-Grain Cars Loaded and Billed: Weekly total number of grain cars loaded and billed, reported by province, identifying totals in shuttle service versus total cars loaded in all other ordering systems

BNSF has populated Data Element No. 7 with total loaded and billed grain cars by province in a manner consistent with the way BNSF reports the number of cars loaded in the CS54 data submitted weekly to the AAR for public reporting. Total loaded and billed cars includes cars in shuttle service, dedicated train service, reservation, lottery, open and other ordering systems, and private cars with active waybills. For purposes of this Data Element, grain includes barley, corn, oats, rye, sorghum, wheat, other grain, soybeans, dry beans, dry peas, lentils, cowpeas and lupines.

• Data Element No. 8-Grain Cars Orders: Weekly and by province, for railroad-owned and leased cars: (a) running total of car orders placed; (b) running total of orders filled; (c) for unfilled orders, number of orders that are 1-10 days past due and 11+ days past due, as measured for when the car was due for placement under the Class I rail carrier's governing tariff

BNSF has collected the data responsive to 49 C.F.R. § 1250.2(a)(8), as referenced by the Act, in a manner consistent with our prior reporting to the STB and our communications to our customers. To identify the running total of orders placed, we report the total number of new car orders placed during the seven day reporting period. The running total of orders filled is the number of that have been placed for loading over the seven day period. Under BNSF's governing tariffs, a pending car order is classified as past due when the shipment is more than three days past the shipper's want date. Any order with a want date that is more than three days old is considered a "Past Due" and any shipment that is Past Due on the Sunday during the current reporting period is allocated either into the 1-10 Day column or the 11+ Day column, depending on its age. Consistent with the Act, the data responsive to this Data Element is limited to traffic moving through Canada.

• Definition of Unit Train: In our periodic report required by the Act, BNSF has maintained our historic train-symbol approach to identifying unit trains, which is used consistently in our company communications and tools used by our customers, and the historic informal and formal reporting provided to the STB, including the reporting required by 49 C.F.R. § 1250.2(a) on which the Act is based. To provide unit train specific metrics across the various reporting requirements, BNSF identifies unit trains by specific train symbols that reflect the specific service offerings available to our customers. For example, coal unit trains are identified by the letter at the start of the train symbol-C for loaded unit trains and E for empty unit trains. BNSF does not include intermodal trains in unit train reporting.
• Reporting Period and Timing: Consistent with s. 77(5) of Act, each of BNSF's reports covers the period of 12:01AM Saturday to 11:59PM Friday with data being presented in the report due no later than five days after the last day of the period to which the information relates.

Conclusion

Consistent with the Act, BNSF will provide an update to the Minister of Transport in the event that BNSF changes the methodology used to generate the periodic report or changes the definition used to identify unit train traffic.

CSXT

Explanation of methodology

Under the Transportation Modernization Act, railways have been requested to report the below eight service metrics for Canadian operations, adopting a subset of current reporting to the U.S. Surface Transportation Board pursuant to 49 CFR 1250.2(a)(1)-(8). The explanations for each category below relate to CSXT’s reporting of these service metrics for its Canadian operations. For further context and clarification on terms and definitions, CSXT adopts as applicable the Explanation of Methodology submitted for CSXT’s regular STB reporting, which can be found on the STB’s website.

1. Train Speed. All CSXT trains in Canada, including cross-border trains, are operated as local trains. Local, passenger, maintenance of way trains and yard jobs are excluded from CSXT train speed measurements. As a result, CSXT anticipates no velocity to report.
2. Terminal Dwell. CSXT has four locations in Canada with reportable dwell events (note: four additional locations have negligible car counts; due to confidentiality concerns the data at these locations have been rolled up into a larger terminal). CSXT will report weekly average terminal dwell time, in hours, for these four locations.
3. Weekly average cars on line. CSXT will report the weekly average of cars on line.
4. Dwell time at origin for unit trains. CSXT anticipates no dwell time at origin for unit trains to report because CSXT does not originate any of the applicable unit train types in Canada.
5. Trains holding. CSXT anticipates no trains holding to report because CSX does not move any of the applicable unit train types in Canada.
6. Weekly average number of cars that have not moved in 48 hours or greater. CSXT will report loaded cars and empty cars in revenue service that dwell for 48 hours or greater. Cars will be reported by the requested classification based on a combination of STCC, car type and CSXT line-of-business. CSXT will report this statistic as the weekly average of seven daily same-time snapshots.
7. Grain cars loaded and billed by state. CSXT anticipates no grain shipments to report because CSXT does not have any grain shipments in Canada.
8. For grain reported in (7), railroad-owned or leased cars that move in manifest service. CSXT anticipates nothing to report because CSXT does not have any grain shipments in Canada. CSXT will periodically validate that there is nothing to report for categories 1, 4, 5, 7 and 8. Reporting will commence for categories 1, 4, and 5 when four or more of an applicable train type are operated during the week. As expressed in CSXT’s Explanation of Methodology to the STB, less than four trains in a given week is not viewed as statistically significant for service metric reporting purposes.

Norfolk Southern

Data Methodology

Overview:

Section 77 of the Transportation Modernization Act (“Act”) requires Class I rail carriers as defined in the Act, which includes Norfolk Southern Railway Company (“NSR”), to report certain service and performance indicators for their operations in Canada until regulations are issued by Transport Canada.

NSR’s only operation within Canada is one train a day that moves over less than two miles of a line of Canadian National Railway Company (“CN”) from the border crossing at Buffalo, New York, into CN’s yard in Fort Erie, Ontario. However, NSR is not the waybill carrier and does not receive any revenue for the Canadian portion of the move. Rather, such movement is purely for operational convenience of the parties to facilitate interchange of traffic between NSR and CN at Buffalo. The train is dispatched at the direction of CN and operated in accordance with CN’s rules and regulations. Consequently, NSR does not currently have any revenue operations or traffic in Canada

As a result, NSR does not have data to report for the interim service and performance indicators contained in the Act. Information on NSR’s system performance in the United States is collected by the Surface Transportation Board and can be accessed via the STB's website.

NSR provides a detailed explanation of the application of each of the service and performance indicators required under the Act to NSR’s Canadian operations below.

1. Average Train Speed.
   NSR calculates train speeds for its U.S. system from data capturing the total train miles in line haul movement as well as the train transit segment move days by train type. NSR’s daily interchange operation with CN is performed with a local train operating out of Buffalo, NY. NSR does not include local trains in its train speed reporting or capture train speed data for local trains. As a result, NSR’s average train speed for all train types in Canada is N/A.
2. Weekly Average Terminal Dwell.
   NSR does not have any terminals in Canada. As a result, NSR’s weekly average terminal dwell in Canada is N/A.
3. Weekly Average Cars on Line.
   NSR does not have any tracks or facilities in Canada. Any NSR traffic moving into Canada is interchanged to CN upon arriving in CN’s Fort Erie Yard, and any NSR traffic received in interchange from CN in CN’s Fort Erie Yard is moved directly across the border and into NSR’s rail system in the United States. Therefore, NSR does not have any cars on line located in Canada. As a result, NSR’s weekly cars on line for all car types is Zero.
4. Weekly Average Dwell Time at Origin.
   NSR does not originate any traffic in Canada. As a result, NSR’s weekly average dwell time at origin is N/A.
5. Weekly Average Number of Trains Holding Per Day and by Cause.
   NSR does not have any tracks or facilities in Canada, and does not have the right to hold trains on the less than two miles of CN track over which it operates to facilitate interchange. As a result, NSR’s weekly average number of trains holding per day is Zero.
6. Weekly Average of Loaded and Empty Cars Not Moving in 48 Hours.
   NSR does not have any tracks or facilities in Canada, and does not have the right to store cars on the less than two miles of CN track over which it operates to facilitate interchange. Cars being interchanged to CN in Ft. Erie go into CN’s account upon delivery, and cars received from CN in interchange in Ft. Erie only enter NSR’s account when NSR picks up the cars to pull to Buffalo. As a result, NSR’s weekly average of loaded and unloaded cars which have not moved in 48 hours is Zero.
7. Weekly Total Number of Grain Cars Loaded and Billed by Province.
   NSR does not originate any traffic in Canada. As a result, NSR’s weekly total number of grain cars loaded and billed in each province is Zero.
8. Grain Car Orders Placed and Filled by Province.
   NSR does not originate any traffic in Canada. As a result, NSR’s weekly total number of grain car orders placed and filled in each province is Zero.

Union Pacific

Background

Union Pacific does not own any rail lines in Canada. Nor does Union Pacific conduct any revenue operations in Canada. Union Pacific’s participation in cross-border rail transportation shipments is limited to the portion of the transportation south of the U.S.-Canada border, with the transportation north of the border provided by connecting railroads operating in Canada. As a result, the revenue Union Pacific receives from shippers is solely for that portion of transportation movements occurring within the United States.

There is one location on the U.S.-Canada border where Union Pacific’s train crews cross a short distance into Canada. At the border crossing between Eastport, Idaho, and Kingsgate, British Columbia, the trains Union Pacific interchanges with Canadian Pacific Railway for southbound movements must be scanned by United States Customs and Border Protection using a Vehicle and Cargo Inspection System (VACIS) machine. This machine uses electronic imaging to scan the train and is located approximately 100 feet south of the border. On those movements, a Canadian Pacific crew stops the train approximately 100 feet north of the border. A Union Pacific crew then moves the train across the border and through the VACIS machine. On northbound train movements, the Union Pacific train crew pulls the train north of the border, beyond the end of Union Pacific’s tracks (which terminate at the border), onto the tracks of Canadian Pacific, and far enough north for the last car on the train to clear the VACIS machine. This maneuver means the Union Pacific train crews travel north into Canada only as far as one train length (approximately one mile) before Canadian Pacific takes over operation of the train.

Methodology

1. Average Train Speed
   Average Train Speed is calculated by dividing train-miles by total hours from origin to destination, less intermediate terminal time. This measure excludes the following train categories: yard, local, passenger, foreign, and maintenance of way.
   For the purposes of reporting under Section 77 of the Transportation Modernization Act (TMA), Union Pacific’s Average Train Speed is either a null quantity or “not applicable”: (1) calculations should not include data from activities occurring in the United States (e.g., the portion of the movement between the U.S. origin/destination and the U.S./Canadian border); (2) the portion of the Union Pacific movement in Canada is limited to the distance sufficient to get the train over the border; (3) Union Pacific crews interchange after the last northbound car clears the border, so there is no true Canadian movement on Union Pacific to a destination; and (4) Union Pacific crews pick up southbound trains approximately 100 feet north of the U.S. border and then proceed through the VACIS machine, thus there is effectively no Canadian movement from origin.
2. Average Terminal Dwell
   Average Terminal Dwell measures the average hours a car resides at the specified terminal location. It begins with train arrival, customer release, or interchange receipt. The measure ends with train departure, customer placement (actual or constructive), interchange offering or delivery. This measure excludes cars moving through the terminal on run-through trains, stored cars, bad ordered cars, and maintenance of way cars.
   For TMA purposes, Union Pacific’s Average Terminal Dwell is a null quantity. Even assuming that the VACIS machine constitutes a “terminal,” the southbound trains are run-through trains – thus no dwell. Even if there were a “dwell”, the measure does not include that “dwell” during a movement by Union Pacific in Canada. For southbound movements, Union Pacific crews pick up the train approximately 100 feet north of the U.S. border and then move the train directly across the border. For southbound traffic, any dwell while on Union Pacific is south of the border, and any dwell in Canada would occur after interchange to Canadian Pacific.
3. Weekly Average Cars
   Average Cars On Line for the Week is the average daily inventory of all freight cars in the revenue fleet regardless of location or status, and includes cars on shortline railroads, cars delivered to customer facilities, and stored cars.
   For TMA purposes, Union Pacific’s Weekly Average Cars is also a null quantity or “not applicable.” Union Pacific has no trackage in Canada, and thus no cars online within Canada. We expect that Union Pacific’s Canadian interchange partners report any cars online in Canada that are delivered by or destined for Union Pacific on cross-border movements.
4. Weekly Average Dwell Times at Origin
   Weekly Average Dwell measures the time from customer release to train departure at origin. The release would be measured using the last cut of five or more cars and would include both loaded and empty movements. This measure excludes trains received in interchange from another railroad and intermodal trains.
   For TMA purposes, this number for Union Pacific is either a null quantity or “not applicable.” Because the statutory measure excludes trains received in interchange from another railroad, southbound cross-border movements received in interchange from Canadian Pacific are, by definition, excluded. For northbound cross-borders movements, any origin on Union Pacific is in the United States, and thus not a Canadian origin.
5. Average Number of Train Holdings (5a Crew, 5b Power, 5c Other)
   Average Weekly Trains Held Short of Destination or Scheduled Interchange is a cumulative measure using daily snapshots of active trains held for more than six consecutive hours and excluding yard and local trains. The measure is broken down by the attributing causes: (a) crew limitations; (b) lack of locomotive power; and (3) other (e.g., track maintenance, mechanical issues). This measure excludes non-Union Pacific holds, including customer holds, holds due to other railroads, rail incidents, and weather.
   For TMA purposes, this measure for Union Pacific is a null quantity. Any holds for southbound cross-border trains are on Canadian railroads, not Union Pacific. Any holds on Union Pacific of northbound cross-border trains are in the United States, not in Canada. Moreover, such holds are excluded from reporting. For example, trains delayed when moving through the VACIS machine, because of customs activity or the lack of track space north of the border, would be excluded non-Union Pacific holds.
6. Average Cars Not Moving (6a Loaded, 6b Empty)
   Average Weekly Total Cars in Service Not Moving is measured using a daily snapshot of in-service freight cars that have not moved for 48 or more hours. The measure starts with the pull from the customer facility or interchange pick-up and excludes cars in hold status, empty cars not billed to a specific consignee, and non-revenue movements.
   For TMA purposes, this measure for Union Pacific is a null quantity. Pulls from customer facilities by Union Pacific occur solely within the United States. For southbound cross-border movements, Union Pacific has no pulls from customer facilities in Canada. Union Pacific’s pick-ups from interchange at the Canadian border involve no cars “not moving” because interchange only occurs when Union Pacific’s crew arrives and actually begins to move the cars across the Canadian border into the United States.
7. Grain Performance
   Weekly Total Grain Cars Loaded and Billed is measured by aggregating the following Standard Transportation Commodity Codes (STCCs): 01131 (barley); 01132 (corn); 01133 (oats); 01135 (rye); 01136 (sorghum grains); 01137 (wheat), 01139 (grain, not elsewhere classified), 01144 (soybeans), 01341 (beans, dry), 01342 (peas, dry), and 01343 (cowpeas, lentils, or lupines).
   For TMA purposes, Union Pacific’s Weekly Total Grain Cars measure is a null quantity. Union Pacific does not load grain cars in Canada, and it does not bill for any transportation provided within Canada. All such billing is by Union Pacific’s Canadian interchange partners.