Top of Rail Friction Control with locomotive delivery on BC Rail: Reductions in fuel and Greenhouse Gas Emissions

Transcription

1 John Cotter 1 Top of Rail Friction Control with locomotive delivery on BC Rail: Reductions in fuel and Greenhouse Gas Emissions Norm Hooper, Jay Roberts and Tony Makowsky, BC Rail, PO Box 8770, Vancouver, BC, Canada John Cotter, Don Eadie and David Elvidge, Kelsan Technologies Corp West 15 th St., North Vancouver, BC Canada Yan Liu, Centre for Surface Transportation Technology, National Research Council of Canada, 2320 Lester Road, Ottawa, Ontario ABSTRACT Top of rail friction control is a new technology offering significant benefits in fuel consumption, rail and wheel wear, and reduced track structure degradation (ties, spikes etc). In 2003, two BC Rail 4400 HP road locomotives (Dash 9-44CW) were outfitted with top of rail (TOR) dispensing equipment. This 12 month project assessed the impact of Kelsan Technologies Top of Rail (TOR) Friction Modifier Technology, KELTRACK, on reducing fuel consumption and Greenhouse Gas (GHG) emissions as well as reducing the state of stress in the track structure. The test utilized a 6000 ton 45 sulphur car consist traveling on BC Rail s Chetwynd subdivision. A total of twenty test trains were run (ten baseline & ten top of rail) between June and November Measurable savings were based on monitoring changes in diesel fuel consumption, as well as mechanical drawbar forces. Analysis of the test data indicated a strong correlation between fuel savings and (increasingly) curved track. In over the road conditions, fuel savings of up to 10 % were observed in heavily curved track (> 50 %). Field testing also successfully demonstrated reduced stress in the track structure in two situations: i) When a loaded consist was

2 John Cotter 2 traveling through a sharp reversing curve (under tractive effort) and ii) when a loaded consist was traveling through a sharp curve on a descending grade (under dynamic braking). No train handling issues (traction or braking effectiveness) were encountered over the course of the trial as monitored by the operating department. This included application of the TOR friction modifier in areas of significant ascending & descending grades (1.2 to 1.5%). Over the course of the 12 month project, the TOR equipment performed to expectations. A stringent field monitoring program during this period resulted in several implemented upgrades to the TOR equipped units in order to further assure reliability in extreme (hot/cold) environments. In one example, the TOR equipment was successfully shown to operate in temperatures down to (-22 F or -30 C) during the winter of INTRODUCTION Top Of Rail Friction Modifier Technology Unlike lubricants, friction modifiers (KELTRACK ) are able to reduce the COF of dry rail and maintain the desired intermediate level of friction over a given number of trains or wheel passes. Friction modifiers formulated for freight applications contain no oils, greases or other liquid lubricant components. They are composed of engineered composite solids, polymers and other compounds that are mixed with water and deposited on top of the rail in liquid form. When the water evaporates, the remaining dry

3 John Cotter 3 thin film (similar to a coat of paint ) maintains an optimal intermediate coefficient of friction. In this application, a water based friction modifier is applied to the Top of Rail behind the last driving wheel of the trailing locomotive. The water-soluble material is applied as a fine atomizing spray to ensure rapid drying and production of a thin uniform film. The water in the material is quickly evaporated under the action of the first few wheels and the dry COF ( ) is reduced to about 0.35 a level that is maintained throughout the length of the train. Since this intermediate level of friction can be maintained, the material need not be dissipated or removed by the last wheel on the train. In fact, the residual friction modifier left on the rail can actually benefit a subsequent train by reducing lateral forces and energy consumption-even if the subsequent train has no friction modifier delivery system of its own onboard. With no need to remove residual material from the rail, compensation for the range of such operating and environmental conditions as speed curvature, creepage, temperature etc is much less critical. As a result, the friction modifier s control system is greatly simplified. TOR Dispensing Equipment The TOR dispensing equipment for the Dash 9-44CW locomotive was designed and installed by Lubriquip, a subsidiary of IDEX Corp. The system is comprised of the following primary components: a control system, product metering pumps, dispensing nozzles, a storage reservoir, and heat tracing to allow the system to nominally operate in temperatures of -22 F (or -30 C).

4 John Cotter 4 The control system has the following features: Curve sensing for differential application on curves Brake actuation sensing to interrupt fluid delivery during braking, Locomotive speed sensing for delivery rate computation A Last Locomotive in Consist (LLIC) algorithm which ensures that the friction modifier is applied only from the last locomotive in consist after all of the driving wheels have passed. Accurate metering of the friction modifier product is accomplished by use of small, highly controllable pumps. The pumps are contained in weather tight enclosures mounted on bulkheads of the locomotive undercarriage. Patented dispensing nozzles, developed by Kelsan Technologies, ensure that an atomized spray of the friction modifier is applied onto the top of rail. The nozzles are self cleaning to eliminate build-up which could lead to clogging. It should be noted that the (2) nozzles on either locomotive never clogged during the entire course of the trial. The nozzles are mounted on the sand tube support brackets, which are located at each end of the locomotive adjacent to the last driving wheels. The friction modifier is stored in a steel powder coated reservoir. The reservoir (working) capacity is approx. 43 US gallons and is located in the locomotive compressor room. These TOR dispensing units were also equipped with a prototype heat tracing system to allow the water-based TOR friction modifier to be applied in low temperatures (-22 F or -30 C). The heat trace system is automatically activated when the ambient temperature drops below 41 F or 5 C.

5 John Cotter 5 Over the course of the 12 month trial, the TOR dispensing equipment was monitored closely. BC Rail instigated an inspection regime where the TOR units were examined every 15 days during the locomotives regular trip inspection. Analysis of the inspection reports, along with visual inspections by the project team, identified areas for potential improvement. Subsequent solutions were first lab tested prior to implementation. The subsequent upgrades were then monitored in the field. Upgrades included: Successfully implementing a more robust solenoid valve for operating in high temperatures (122 F or 50 C) Successfully implementing the use of cartridge heaters in dispensing nozzle body for operating in low temperatures (-22 F or -30 C). Modifying the control software to allow for more economical application of the TOR friction modifier when the consist is under dynamic braking. Test Locale Testing was conducted on BC Rail s Chetwynd subdivision between Chetwynd and Prince George, British Columbia. The terrain is primarily rolling hills with several sections of ascending & descending grades (figure 1).

6 John Cotter Segment 2 Segment 3 Segment Segment Segment 1 Segment 5 Segment 7 Segment Elevation (ft) Direction of Travel Segment Mile Board Figure 1: Division of test track into segments. For test purposes, the track was divided into nine segments based primarily on grade. The percentage breakdown between curve and tangent track was also determined for each segment. Further statistics for each segment may be found in table 1.

7 John Cotter 7 Segment # Distance Elevation # Curves % Curves (Distances) Comments 1 12 miles Elevation climb from < 5 : % Continual uphill 695 ft to 758 ft < 5 < 8 : 7 climb > 8 : miles Elevation climb from < 5 : % Steep Uphill 758 ft to 876 ft < 5 < 8 : 11 > 8 : miles Elevation change from 876 ft to 879 ft 4 10 miles Elevation change from 879 ft to 734 ft 5 25 miles Elevation change from 734 ft to 733 ft 6 39 miles Elevation change from 733 ft to 715 ft 7 22 miles Elevation change from 715 ft to 722 ft 8 16 miles Elevation change from 715 ft to 589 ft 9 11 miles Elevation change from 589 ft to 574 ft < 5 : 8 < 5 < 8 : 5 > 8 : 0 < 5 : 22 < 5 < 8 : 13 > 8 : 10 < 5 : 46 < 5 < 8 : 13 > 8 : 4 < 5 : 61 < 5 < 8 : 13 > 8 : 1 < 5 : 14 < 5 < 8 : 6 > 8 : 0 < 5 : 8 < 5 < 8 : 1 > 8 : 0 < 5 : 14 < 5 < 8 : 4 > 8 : 1 Table 1: Test Zone 51 % Flat 47 % Downhill 36 % Rolling Terrain 34 % Rolling Terrain 25 % Rolling Terrain 15 % Downhill 35% Flat TEST METHODOLOGY- FUEL CONSUMPTION & MECHANICAL WORK The test train consisted of two Dash HP locomotives with 45 loaded sulphur cars (trailing tonnage of approximately 6000 tons). The motive power remained unchanged for the duration of the test. The TOR dispensing equipment was located in the trailing locomotive, BCOL Specialized data acquisition equipment was also installed onboard BCOL 4643, to monitor: i) Pertinent locomotive operating parameters; namely traction motor currents, main generator voltage, manifold temperature, barometric pressure, and parasitic auxiliary loads. ii) Drawbar forces through the use of an instrumented shear pin.

8 John Cotter 8 iii) iv) Test train location via an onboard GPS unit. Critical TOR system dispensing parameters. The project consisted of two phases, each containing approximately five baseline runs and five TOR application runs for a total of twenty test runs. All data captured was sorted by Kelsan Technologies and then packaged for distribution to the National Research Council Centre for Surface Transportation Technology (NRC-CSTT) for complete data analysis to derive fuel savings and corollary green house gas reduction. BC Rail was responsible for marshalling the forty-five car unit sulphur trains and scheduling the test runs from the sulphur load-out site and/or nearby staging tracks and operating the test trains to Prince George. Mechanical work & fuel consumption were calculated in the following manner for each test segment: Mechanical Work: i) Derive raw mechanical work from drawbar force & train speed for each segment (excluding incidents of dynamic & air braking). ii) Correct raw mechanical work (i) for any changes in train momentum (difference between final and initial train speed for each of the run segments). This factor compensates for any variation in handling due to different train crews. iii) Corrected mechanical work is then divided by the ton-miles associated with the segment.

9 John Cotter 9 Fuel Consumption Method A: i) Derive discrete fuel consumption data for each change in notch setting (excluding incidents of dynamic & air braking). ii) iii) Correct fuel data (i) for changes in temperature and barometric pressure. Correct fuel data (ii) to exclude fuel consumption used for auxiliary or parasitic loads. iv) Sum total fuel consumption (iii) for each segment. v) Correct 1 total fuel consumed (iv) for any changes in train momentum. vi) Total fuel consumed is then divided by the ton-miles associated with the segment. Note: 1 Unlike mechanical work, accounting for changes for momentum on a fuel basis is more difficult. A correlation factor was developed to establish a relationship between fuel consumption and mechanical work data for the test consist. The correlation factor would build in all of the unknown parameters such as engine(s) efficiency, fuel BTU value etc. The correlation factor would then be used to convert the mechanical work associated with any changes in train momentum to a fuel value. It is important to note that the conversion data is based only on data that has been filtered for dynamic & air braking.

10 John Cotter 10 Method B The methodology is similar to Method A except that the correlation factor is applied to the corrected mechanical work value in order to derive the corrected fuel consumption value. Method A utilizes discrete fuel data as derived from individual notch settings. However, in incidents when the notch residence time is small, the corresponding fuel consumption rate used for data calculations may not be accurate as changes in fuel rates are not instantaneous. On the other hand, the correlation factor utilizes a composite fuel value based on a wide range of notch settings for a given amount of mechanical work. Both methods were used in the calculations and the data is presented for perusal by the reader. It should be noted that for either method, the focus was on determining the difference between baseline & TOR conditioned rail rather than on ascertaining the absolute fuel consumption values. FIELD TESTING METHODOLOGY TRACK STRUCTURE Scenario #1: Sharp Reversing Curve under Tractive Effort Many field tests have been performed which clearly demonstrate the ability of TOR friction modifier technology to reduce the state of stress in the track structure in a single curve. For demonstration purposes, a severe scenario was assessed in order to determine how TOR friction modifier technology can improve truck steerability in a sharp reversing curve.

11 John Cotter 11 The main criteria used in selecting the field test site were the presence of a sharp reversing curve (curvature above 8 (218 m)), wooden ties and standard cut spike fastening. After reviewing BC Rail s track profile and operating personnel, a suitable site located at mile 598 meeting track requirements with suitable track access was selected (figure 2). The test consist would be under tractive effort at this site due to the presence of an ascending grade. Rail deflection gauges (figure 3) were placed in the full body of the curves 1 & 2 as well as in the middle of the small tangent section. Both (low/high) rails were outfitted with deflection gauges. The top of rail CoF was also monitored before and after the TOR applying test train using a Salient Systems tribometer. Bracket Tangent Curve #1 Curve #2 Base Plate Figure 2 : Test Section Figure 3: Rail Deflection Gauge Scenario #2: Single sharp curve under Dynamic Braking An additional scenario was also investigated; specifically, the impact of TOR friction modifier technology on reducing lateral forces while a consist is under dynamic braking.

12 John Cotter 12 Under dynamic braking, no fuel savings are achievable when applying a friction modifier. In fact, braking requirements may increase slightly due to the reduction in curving resistance of the (loaded) trailing cars. However, application of the TOR friction modifier on descending grades can help to reduce lateral forces in sharp curves. Experiments focused on optimizing TOR application rates/strategies in order to reduce lateral loads for consists under dynamic braking. BC Rail maintains an instrumented 12 o (146 m) curve near Clinton (MP 212) to measure lateral & vertical loads. This site, for loaded southbound trains, is on a descending grade that results in the use of dynamic braking (as well as occasionally, air braking). Test consists usually comprised of 3 locomotives for the motive power (trailing TOR equipped) with approximately 100 trailing cars (approx. 10,000 tons). Since these consists are comprised of mixed freight, data analysis involves screening the consist for a minimum vertical load to ensure a reasonable comparison of baseline and TOR applying trains. Further filtering is also performed to remove trailing axles from the data which typically have low lateral force values. Focus of analysis was usually on the high rail forces.

13 John Cotter 13 MECHANICAL WORK & FUEL SAVINGS RESULTS Mechanical Work Table 2 highlights the results from the mechanical work calculations. Segment Note 1:MTM- Million Ton-Miles % Curve Density Mechanical Work Savings (KWHr/MTM 1 ) % Savings 1 34 % % 2 42 % % 3 51 % 2, % 4 47 % Descending Grade 5 36 % 1, % 6 34 % % 7 25 % % 8 15 % Descending Grade 9 35 % % Table 2: Mechanical Work Savings In reviewing the tabulated results, there does appear to be a general relationship between increasing mechanical work savings and increasing track curvature (except for segment 1). Segment 3, for example, observed a 12 % reduction in mechanical work in a heavily curved (51 %) section of track. The low reported savings for segment 1 can be attributed to the fact that the TOR system was turned on immediately at the start of this segment. As a result, the low value is likely due to insufficient time for the friction modifier to work through the entire length of the test consist. Fuel Savings Fuel savings, calculated by both methods, are presented in table 3. Method A calculations were performed by the NRC. Method B data analysis was performed by Kelsan Technologies based on the mechanical work calculations provided by the NRC.

14 John Cotter 14 Figure 5 provides an example of the excellent correlation between mechanical work & fuel consumption. Segment Curve Density Method A Method B Average % % Savings Savings Fuel Savings (USG/MTM) Fuel Savings (USG/MTM) Fuel Savings (USG/MTM) % Savings 1 34 % % % % 2 42 % % % % 3 51 % % % % 4 47 % Descending Grade 5 36 % % % % 6 34 % % % % 7 25 % % % % 8 15 % Descending Grade 9 35 % % % % Table 3: Calculated Fuel Savings In reviewing the data, there again appears to be a strong relationship between fuel savings and track curvature for both models which follows the same trend established with the mechanical work data. Some discrepancy was observed for data generated from segment 5 (36 % curve density). Figure 4 shows the relationship when the results from the two methods are averaged. Fuel Savings US Gallons/MTM % Curve Density Figure 4: Fuel Savings

15 John Cotter 15 Fuel Comsuption (Liter) Test Data of base03_1.dat Linear (Test Data of base03_1.dat) y = x Mechanical Work on Trailing Tonnage (kw-hr) Figure 5: A typical example of relation between fuel consumption and mechanical work FIELD TESTING RESULTS TRACK STRUCTURE Scenario #1: Sharp Reversing Curve under Tractive Effort Figure 6 highlights the ability of the TOR friction modifier product to control friction between 0.3 & 0.4 after the TOR applying train traversed the test zone. Notice the large variations in the measured top of rail friction associated with the baseline conditions. GHG TOR Test Train #3 - Friction Measurements July 16, 2003 Coefficient of Friction Test Curve #1 Test Curve #2 High Rail Low Rail Low Rail High Rail Tunnel Test Curve Section (100 ft intervals) West Rail before West Rail after East Rail before East Rail after Figure 6: Measured Top Of Rail Coefficient of Friction

16 John Cotter 16 Rail Deflection Gauge (RDG) data for the second curve are shown in figures 7 (baseline) & 8 (TOR applying). The speed of the baseline & TOR test train was approximately mph in each case Positive Deflection is West to East "0" deflection - static rail conditions (no train) 7.0 Rail Deflection (mm) Train Axle No. Low Rail High Rail 2 per. Mov. Avg. (Low Rail ) 2 per. Mov. Avg. (High Rail ) Figure 7 - Baseline Train RDG Data Curve # Positive Deflection is West to East "0" deflection - static rail conditions (no train) 7.0 Rail Deflection (mm) Train Axle No. High Rail Low Rail 2 per. Mov. Avg. (Low Rail ) 2 per. Mov. Avg. (High Rail ) Figure 8 - TOR Train RDG Data Curve #2

17 John Cotter 17 Comparing low rail deflection results from curve #2 (Figures 7 & 8); the rail deflection and its variability were both significantly reduced with TOR application. Although the TOR application high rail deflection was slightly increased from 2.62 mm to 3.20 mm, the deflection variability (standard deviation) was also reduced over the baseline condition. The rail deflection data collected supports the findings that TOR treatment improved truck steering through the reversing curve #2. By providing uniform low and high rail friction conditions, improvements in truck steering were evident by the reduction in low rail deflection. The higher low rail deflection and variability during baseline conditions were attributed to increased truck warping that was reduced through the application of the TOR friction modifier. Rail Deflection Gauge (RDG) data for the short tangent section of track is shown in figures 9 (baseline) & 10 (TOR applying). The speed of the baseline & TOR test train was approximately mph in each case.

18 John Cotter West Rail - Positive deflection is East to West East Rail - Positive deflection is West to East "0" deflection - static rail conditions (no train) 3.0 Rail Deflection (mm) Train Axle No. West Rail East Rail 2 per. Mov. Avg. (East Rail ) 2 per. Mov. Avg. (West Rail) Figure 9 Baseline Train RDG Data Short Tangent Section West Rail - Positive deflection is East to West East Rail - Positive deflection is West to East "0" deflection - static rail conditions (no train) 3.0 Rail Deflection (mm) Train Axle No. West Rail East Rail 2 per. Mov. Avg. (East Rail ) 2 per. Mov. Avg. (West Rail ) Figure 10 TOR Train RDG Data Short Tangent Section Analysis of the data from figures 9 & 10 also highlights an improvement in the steerability of the trailing trucks as they negotiate the tangent section between the sharp reverse curves.

19 0 John Cotter 19 Scenario #2: Single sharp curve under Dynamic Braking Figure 11 provides an example of the data analysis performed on the southbound loaded consists traveling through the instrumented curve. The figure shows a total of 8 southbound trains which traversed through the site over a 2 day period. All the trains have been filtered to remove low weight axle loads. Reviewing the data, the three baseline trains immediately preceding the TOR train ( ), appears to have high rail lateral forces in the range of kips Lateral Load Vertical Load Lateral Load Vertical Load Figure 11: Example of lateral force reduction for TOR applying train under dynamic braking The lateral force data associated with the TOR applying train (Red arrow) showed a definite decrease in lateral forces. In this case, significant lateral load reduction was achieved by applying the TOR product only in the curve. Further work is ongoing to optimize the application rate/strategy. As a result of this work, the control system

20 John Cotter 20 software has been upgraded to allow the user to use an alternative application rate/strategy when the consist is under dynamic braking. TRAIN HANDLING During the course of this project no train handling issues were reported by the train crews or even observed by the crew when interviewed on trips when the TOR friction modifier was applied. This includes: When the TOR product was applied during dynamic braking on descending grades to reduce lateral forces in curves. For trains following the TOR test train When air braking was used to decelerate a train. It should be noted that application of the TOR friction modifier was inhibited during activation of the air brakes. Some changes in train handling behavior were observed when analyzing the data; this included longer time/distances spent in dynamic braking (figures 12 & 13) as well as additional use of air braking to assist dynamic braking on descending grades if the TOR product is applied.

21 John Cotter 21 Dynamic Braking Time (Seconds) Segment 3 Segment 5 Segment 6 Segment 7 Baseline TOR Figure 12: Segment Dynamic Braking Time (Seconds) Dynamic Braking Distance (Miles) Segment 3 Segment 5 Segment 6 Segment 7 Baseline TOR Figure 13: Segment Dynamic Braking Distance (Miles) CONCLUSION The project has successfully demonstrated that TOR friction modifier technology can: Provide significant fuel savings in areas of high track curvature. Savings were also validated by reductions in mechanical work as measured by an instrumented shear pin. Provide significant reductions in the state of stress in the track structure; especially in sharp reversing curves. Does not impact train handling. Successfully operate in temperatures down to (-22 F or -30 C).