Incentivized decarbonization through safer and more efficient heavy haul operations

Freight Railway Operations

The Canadian freight railway network spans over 42,000 km of track. Freight trains accumulate over 100 million train-km carrying more than 300 million carloads annually. ¹³ CP and CN, whose networks are shown in Figure 1, handle over 95% of the Canadian annual tonne-kms and operate more than 75% of the railway tracks. ¹⁴ OPEN IN VIEWER

The transport sector is the second largest source of greenhouse gas (GHG) emissions in Canada. In 2019, transport sector operations emitted a total of 185.5 Mt-CO₂e. ¹⁵ Of which, 13.1 Mt-CO₂e was emitted from freight transportation by air, marine, and railway. Although this number is low compared to Canada’s overall GHG emissions, reducing emissions wherever possible, even in systems which are already efficient, will still be important. There are hard-to-abate emissions that are difficult or entirely infeasible to eliminate, ¹⁶ but in the case of the railway industry, there is potential to further reduce emissions in areas which are not only possible but also provide financial incentives to the operators and infrastructure owners. Most of the Canadian railway network has a single main track. Delays experienced by one train can have significant downstream effects and increase unnecessary emissions.

Railway Equipment and Track

Canadian railways operate approximately 62,000 cars and 3800 locomotives. Many of these cars and locomotives are regularly interchanged with railways operating in the US and Mexico. Standardization in railway equipment, especially in cars, has enabled seamless traffic interchange.

Given the overwhelming number of four-axle cars compared to locomotives and other types of freight cars, the discussion in this paper will primarily be based on the conventional four-axle car. With some exceptions, almost all of the car’s mass is from carbon steel. General purpose cars typically have a total (gross) loaded mass of 130 t, and bulk commodity cars can have a total loaded mass of up to 143 t. The tare mass ranges from the order of 27–34 t, and the (net) mass of the commodity that the car is carrying is more than 3 times the (tare) mass of the car itself.

The construction of the track requires moving significant amounts of earth and rocks to form the roadbed and ballast. In addition, large structures of various construction materials, signal wires, fencing, culverts, and tunnels may be required depending on the geography of the railway line. Embedded in the ballast lays ties made with timber or concrete, which the rails are fastened to through fasteners and tie plates. Standard North American rail mass ranges from 56.7 to 69.8 kg/m, and the most common mainline rail is 67.4 kg/m. ¹⁷

Much research has been done on the emissions associated with urban and high-speed rail, ¹⁸ but there is a lack of data when it comes to estimating the embodied carbon emissions associated with North American freight railways. Recent estimates suggested that the construction of ballasted track emits 4.7 t-CO₂e/km. ¹⁹ Although this figure included catenary wires which would not be the case on Canadian freight railways, the distance over which raw materials are shipped for construction in the UK would also be much shorter.

Data from the steel and automotive industries were examined to estimate the emission from rollingstock manufacturing. Steel production is a carbon-intensive process. The steel industry estimates that their production makes up 7%–9% of global GHG emissions. ²⁰ Even with the most current methods, every tonne of steel produced leads to 1.85 t-CO₂ emissions. It is also estimated that each automobile produced in the EU leads to 0.61 t-CO₂e. ²¹ Assuming an average mass of 1.5 t for an automobile, the emissions associated with the production-to-raw-material ratio is 0.41 t-CO₂e/t. Using the same ratio and acknowledging the much more diverse use of materials and processes in automotive manufacturing, the emissions from the manufacturing of a freight car can range from 12.3 to 55.5 t-CO₂e using a nominal tare of 30 t.

The Link to Decarbonization

Improving the reliability and service life of assets has always been a focus for railway companies. The expense of replacing rollingstock and track provides a great financial incentive for companies to innovate and develop better materials and practices. Based on the emissions from the construction of components and railway infrastructure, there is also a clear environmental incentive to prolong equipment life and reduce service interruptions.

The most cost-intensive railway components are the wheelsets and rails, which require frequent replacements. The case for longer wheel and rail life is an obvious one, both financially and environmentally.

From the emissions perspective, the incentive for reducing service interruptions is less obvious. The term service interruption encompasses all delays, including unplanned train stops caused by reasons internal and external to the railway system. Especially on a single-track railway network, any delay experienced by a train can have a significant downstream impact. Common delay causes related to car components include overheated bearings, wheels (due to sticking brake), and high-impact wheels, all of which could result in unplanned train stops so that the defective cars could be removed from the train. In addition to the extra fuel burn for the affected train, idle time is increased for the following and opposing trains, leading to further increases in emissions.

In the most severe case, derailments lead to the need to replace both rollingstock and track, causing significant delays. A reduction in derailment would, therefore, reduce both cost and emissions.

Longer Life and Fewer Stops

Prolonging the service life of railcar components has the mutual benefits of lowering costs to the railway and reducing carbon emissions. Some of the most visible efforts in attempting to prolong car component life lie in the development of better steel and maintenance practices.

It is commonly accepted that wheel and rail life does not extend from increasing material hardness alone. In many cases, wear helps remove the microcracks forming at the material surface, preventing them from propagating and creating more severe defects. ²² The material strength and hardness and maintenance practices need to strike a balance in optimizing the life of railway assets. With the right maintenance, it has been shown that stronger steel leads to more satisfactory results in wheel and rail life. Advances in maintenance practices have included optimization of transverse wheel and rail profiles, advances in rail grinding, and more recently milling, to restore target profiles and remove surface damage including rolling contact fatigue.^23–26 Another major development in this respect has been the widespread use of wheel-rail friction management, including both gauge face (GF)/wheel flange lubrication and top-of-rail (TOR)/wheel tread friction modification.^27–30

Higher-performance rail and wheel materials have been evaluated over the years in the test environment and revenue service.^31–34 The ranges of hardness requirements for conventional and higher-performance wheel and rail steel are summarized in Table 1.

OPEN IN VIEWER

Table 1.

Surface hardness requirements for wheels ³⁵ and rails. ¹⁷

	Brinell hardness [HB]
	Standard	Intermediate performance	High performance
Wheel	302 to 341	321 to 363	341 to 415
Rail	310 minimum	325 minimum	370 minimum

Wheel Service Life Improvement

North American freight railways use wrought and cast carbon steel wheels. For cars, the Association of American Railroads (AAR) Class B, C or D wheels must be used. ³⁵

From 2016 to 2022 YTD, over 30,000 wheelsets were replaced by Golden, BC, the primary shop for CP’s coal fleet running from the mines in the Rockies to the port of Vancouver. More than 50% were due to high-impact wheel defects, which occur when the running surface or tread of the wheel was shelled excessively out of round. Both scenarios result in the increase of vertical force applied to the rail from the wheel (over 400 kN from a single wheel), which increases fatigue to infrastructure as well as the risk of fracture. Heavy and prolonged brake application due to mountain grades is a major contributing factor to wheel shelling, an example of which is shown in Figure 2.OPEN IN VIEWER

Class C wheels are the most common for Canadian cars today. Class D wheels were designed to have superior resistance to tread damage without compromising other performance characteristics. ³⁵ A field study to evaluate and compare the performance of Class C and D wheels ran from 2013 to 2018 in the CP coal fleet. During the study, data on wheel survival rate, wheel profile measurements, failure mode, surface condition and mileage were collated. The Class D wheels were found to have higher wear and fracture resistance when compared to their Class C counterparts. By 2018, approximately 60% of the Class C wheels had failed due to high impact, compared to less than 10% of the Class D wheels. High flange was the main cause of Class D wheel removal. The study showed that Class D outlived Class C by over twice as long, with the 50% rejection life of the Class D wheels at 966,000 km versus Class C at 418,000 km. The total number of wheelset removals in Golden as the adoption of Class D wheels increases is shown in Figure 3.OPEN IN VIEWER

Improvements in truck design has historically played a role in the decreasing trend in wheel replacements. ³⁶ The majority of CP’s coal fleet was equipped with trucks with similar performance to the latest performance standard introduced in 2002. ³⁷ There has been no major update to the standard, and maintenance practices can have a significant impact on the in-service truck performance, ³⁸ improvements in both truck design and maintenance can lead to further improvements in wheel life.

Rail Service Life Improvement

Rails on curves typically experience much higher wear than on tangent tracks due to flange contact. Both the gauge (side) and head (vertical) wear are present on the high rail, while the wear on the low rail is primarily head wear. Rail replacement programs at CP are predominantly targeted in curves and are referred to as “curve patching.” Rail programs are capital intensive with more than half of the cost in steel materials in the form of rails and tie plates. Ideal track renewal cycles include the replacement of rails on tangent tracks every 30 years or longer and curve patching every 2–15 years, depending on curvature.

Since the early 2000s, CP has developed a research-based process of friction management combined with regular rail grinding in Western Canada, where the majority of sharp curves are located. ³⁹ The key implementations include:

• Large-scale adoption of wayside friction modifiers to ensure the effective application and control of GF and TOR lubrication.

On curves, the implementation of the frictional management system saw a fuel saving of approximately 40% and a decrease of head wear by 50% and virtually eliminated low rail gauge wear compared to prior data. As a result of continuing to adopt premium rails, the system rail replacement track-km continued to trend down, as shown in Figure 3.

One of the concerns that are often raised amongst railway practitioners and researchers is that an increase in the hardness of one partner in the wheel-rail system will lead directly to an increase in the wear of the other. Recent meta-studies of research in this area have indicated that while care does need to be taken in maintaining profiles as harness increases, there is not a direct or causal relationship that links an increase in the hardness of one partner to an increase in wear of the other.^40,41 While a causal study of the uptick in rail replacement in 2020 has not been done, aging infrastructure, improving inspection techniques, and shortened inspection periods can all contribute to an increase in rail replacement.

Rolling Stock Design

Cars build since the 1970s have a life of 50 years in interchange service, ⁴² and their design continues to evolve throughout the years to become more efficient. Examples of these improvements can be shown using coal and grain cars. Together, coal and grain represent approximately 10% of Canadian railway freight by tonnage. ⁴³

Coal car bodies have a bathtub-shaped design to maximize space efficiency and, therefore, capacity. Aluminum cars are rare in the general fleet, but newer coal cars have been built exclusively with aluminum car bodies, which lead to a drastic reduction in tare mass.

Most grain hoppers are still of steel construction, however, as design and manufacturing techniques improved, their capacity-to-tare ratio has also improved. Newer grain cars are shorter, with higher capacity, so more cars can be processed within the same physical confines of the existing facilities. The current CP grain hopper fleet consists of cars with three distinctive designs from the 1970s, 1990s, and 2010s. The car design has become more efficient with each iteration. As shown in Figure 4, compared to older designs, aluminum coal cars now carry 28% more mass per unit of car mass and grain cars 17% more.OPEN IN VIEWER

Technology

Most of the Canadian railway network is on a single track through remote territories. A delay or service interruption of one train en route will cascade, affecting other trains, work blocks, equipment and staffing for the area. Several technologies are used to reduce service interruptions and serious incidents. Primarily, an expansive automated wayside detector network is used to capture relevant train data en route, improving the safety and efficiency of operations.

Undetected bearing and wheel defects can lead to catastrophic consequences such as multi-car derailments. For this reason, much of the development in wayside technology has been directed at capturing faults in these components. Examples include bearing and wheel temperature, wheel impact force, bearing acoustics, and visual inspections using camera-based systems. Data from the wayside network are fed into a central database and safety and mechanical defects are subsequently detected using proprietary algorithms.

Rolling Stock

Two examples of rollingstock improvements have been discussed, namely, more efficient car design and longer wheel life. While the overall coal cars remained similar since the introduction of aluminum design in the early 2000s, the number of wheelset replacements at CP trended downward as more Class D wheels were used in unit coal train service.

Data from Figure 3 showed that compared to the height of 6000 wheelset replacements in 2017, 2021 saw a reduction of approximately 2600 wheelsets. They represent approximately 2.4 kt-CO₂e of embodied emissions from the casting process, ⁴⁷ based on an estimated wheelset mass of 1400 kg.

Track

There has been a moderate reduction in rail replacement on CP in recent years. Compared to 2017, 93 fewer track-km were replaced in 2021. Unlike newly constructed tracks, the average emissions for track maintenance from data compiled by Krezo et al. are 1.2 t-CO₂e/1000 km. ⁴⁸ As is with wheels, using an average of embodied emissions from the casting process of 0.65 t-CO₂e/t. ⁴⁷ The replacement of 93 km of tracks using 67.4 kg/m rails leads to an estimated 8.1 kt-CO₂e emissions.

Service Interruption

Fuel consumption data from CP unit coal trains in April 2022 from three subdivisions were analyzed. A total of 661 min of delays were accrued, resulting in additional fuel consumption of 1200 L.

At the same rate, an additional 14,400 L of diesel fuel can be consumed over a year which is solely due to service interruptions. Using an estimate of 2.7 kg-CO₂/L from diesel engine emissions, a total of 39 t-CO₂ reduction can be further achieved from unit coal trains alone on CP if service interruptions are eliminated.

Derailment Trends

Derailment is the most severe form of service interruption and can cause major damage and cascading delays to many trains. The reduction of derailments in both number and severity would impact the railway’s emissions significantly. Over the last almost two decades, the railway industry in Canada successfully brought the number of mainline derailments from the order of 200 per year to less than 100, while the volume of freight increased.^49,50 The derailment and freight volume data are shown in Figure 6. From the replacement of derailed rollingstock alone, a reduction from 4000 to 1000 per year represents a reduction of over 6.7 Mt-CO₂e using the emissions estimates presented. Significantly more emissions would also be produced in association with the replacement of the lost shipment and their transportation. The achievement of zero rollingstock loss from derailment represents a potential further reduction of 2.2 Mt-CO₂e per year.OPEN IN VIEWER

From 2019 to 2022 YTD, mainline derailments across the CP network contributed to over 2100 h of system-wide delays. These delays represent an average of 540 h/year. Using the average fuel consumption for coal trains, which is 1.82 L/min, mainline derailments at their current level cause an increase in diesel consumption of 59,000 L. From locomotive emissions alone, 159 t-CO₂ per year in potential emissions could be eliminated by further reducing the occurrence of mainline derailments.

Limitations of the Analysis

Many CO₂e calculations presented are based on data from European research, mainly due to the lack of North American data on embodied emissions. However, railway companies do offer shippers calculators for their carbon emissions from transportation, although the calculation methods and granularity of input data required differ significantly.^51–53 Assumptions have also been made using train services where data are readily available but are generally applicable to railway operations with similar geographical and traffic patterns to the train services presented. It is important to acknowledge that there are limitations in validity when extrapolating to other railway systems, especially those whose infrastructure, equipment, and management systems differ significantly.

The manufacturing of interchangeable freight railway equipment is entirely domestic in Canada, US, and Mexico. As a result, the emissions associated with transporting finished equipment to the owner are neglected in the analysis presented here. Since freight cars typically do not travel loaded in both directions, the emissions associated with the transportation of equipment would only contribute to a minor percentage of the emissions during their service life.

Ongoing Challenges

Some of the biggest challenges in further improvement in wheel and rail life are around winter operations and the increasingly large seasonal temperature fluctuations. Recent years saw temperature swings from −40 to 40°C between winter and summer, making the operating environment harsher for both equipment and infrastructure.

Data quality presents another challenge in the accurate recording of incidents. When service interruption data are not recorded by automated wayside detectors, the information is often handwritten first, and errors can easily occur when transcribed into a database. Wayside detector data are more reliable but are not without problems. Wayside detection is reliant on train matching, which can be caused by late or inconsistent train consist entries. False readings and car-mismatching could mislead data algorithms and fail to prevent potential service interruptions. While onboard monitoring can overcome the limits on the effectiveness of wayside systems due to their finite number and discrete locations, the lack of a power source on freight cars and complexities in practical implementation present significant challenges, that onboard systems will have to overcome.

Railways continue to operate longer trains to optimize crew and asset utilization. Operating mixed-long trains pose unique challenges due to the difference in weight and length between cars on the train. The placement of remote-controlled locomotives and cars becomes crucial to control in-train forces and minimize derailment risk. As train length continues to grow, existing train marshalling rules may become less effective at preventing train make-up-related accidents. ⁵⁴

With wider adoption of technology and further research, railways can overcome the challenges and achieve further reduction in service interruption, associated costs, and environmental impact.