The Pacific Northwest Rail Corridor incremental plan (for the Portland OR - Vancouver BC section) was developed jointly by Burlington Northern Railroad and Wilbur Smith & Associates, which was subsequently joined by Morrison Knutsen International and HDR Engineering. Full development of the initial plan took about 3 1/2 years. Much of the line is single track. A part of the line is double track current of traffic and the rest is two main track CTC. The territory includes six congested terminal areas.
Several assumptions were made as the plan was developed:
Part of the project area had Amtrak service. Amtrak service had been discontinued on part of the project area. The first phase of development was a short-term plan to improve the existing service and restore the discontinued service.
Improvement of the existing service was accomplished by reducing the running time for passenger trains on the line. There were numerous municipal speed restrictions along the line. Each was eliminated through a process that involved the cooperation of the funding agency, the regulatory agency, the railroad and Amtrak. Some of the restrictions had previously been eased only for Amtrak trains through a process involving only Amtrak. The process associated with the project was directed at elimination of restrictions for freight as well as passenger service. The explanation used at hearings included Federal preemption, as track condition and speed is regulated by FRA, as well as the need to minimize the speed differential between trains as much as possible to reduce the possibility that the public would assume that the approaching train was moving slowly. Reduction of the speed differential has the added benefit of increased capacity. The elimination of some restrictions was contingent on such safety measures as fencing, barriers between the track and closely adjoining roadways, and traffic signal improvements at intersections adjacent to crossings.
Concurrent with the process to eliminate the municipal speed restrictions, public crossings, with a few exceptions, were equipped with automatic signals and gates with constant warning detection. Crossings already equipped with automatic signals and gates were upgraded to constant warning equipment. The detection speed for all crossings was set to 79 mph regardless of track geometry restrictions, anticipating yet undecided tilt train equipment. Curve superelevation was adjusted to achieve the maximum possible speed for conventional passenger equipment, assuming the acceleration of conventional equipment.
This first set of improvements yielded a 5 minute reduction in passenger train running time, thus providing the needed visible improvement resulting from the expenditure.
A second part of this phase developed a program of eliminating speed restrictions, or increasing the restriction speed, related to local conditions. This program included improvements such as CTC in lieu of yard limits/ABS, improved bridge/rail locks on drawbridges and improved drainage at problem areas. In conjunction with these changes, Washington State Department of Transportation and Amtrak decided upon and purchased Talgo tilt train equipment. New speed limits were established for this equipment, making best use of the tilt capabilities as well as the faster acceleration and braking of this much lighter equipment. Speed limits were rounded to the nearest one mph instead of the nearest five mph below. This, combined with the faster acceleration and braking increased the number of curve speed restrictions but reduced the impact of the restrictions. For example, a two mile zone of 70 mph for Superliner equipment might be two short zones of 74 mph separated by a 77 mph zone. Because of the large number of curves, this method of establishing speed limit zones reduced the running time by about six minutes over the next lower five mph method. One additional round trip train was added to the existing service in conjunction with these changes. The additional schedule was designed to best serve the commercial requirement of the passenger service, yet operate to the extent possible during the least congested period of time. The ability to operate the additional pair of schedules at the chosen time was approved by train dispatchers and operating officers without the use of simulation testing.
The new service, mostly on a single track line, involved developing a schedule that fit, to the extent possible, with the existing freight operation in order to limit capital expenditure to the available amount. Freight service on the line was already structured, to a degree, because of the distance between sidings and the length of sidings and yard tracks. As with the improved existing service, a program of municipal speed restriction elimination, track condition speed restriction elimination and automatic grade crossing signal installation was undertaken. New running times were determined using a Train Performance Calculator. A simple stringline model was used for planning the schedule and determining the required capital projects, which included new storage yard tracks to allow use of existing sidings for meeting and passing trains. A similar process was used to determine the capital projects for the second pair of passenger schedules on this line. The effect of the changes developed using this method was checked using the Dispatch Planning Model (Berkeley Simulation Software).
The initial feasibility study generated the commercially required running time and service frequency. All subsequent activity was directed at achieving that goal. The long term approach was to determine what changes would be required for the proposed traffic speed and density without regard to division of benefit and responsibility, leaving those considerations to subsequent negotiation. Thus the development could maintain focus on what was necessary to accomplish the desired result. This process involved tools ranging from simple stringline and spreadsheet analysis of schedules, running time, and yard capacity to simulation of proposed solutions with Dispatch Planning Model. Specific signaling solutions beyond the need for CTC, the ability to operate at relatively short headway, and the need for some type of cab signal/automatic train stop to allow high speed were not identified. The long term result was simulated assuming that separate analysis would generate the signal system required to provide the desired transportation result.
The first step was to identify solutions for existing congestion that limited any increased passenger service or caused reliability problems for existing service. These areas would require correction regardless of the long term goal. This involved analysis of the activities generating the congestion such as trains stopping on main tracks to set out, pick up or switch, trains bunching because of congestion at other locations, and crew changes. Generally, these solutions turned out to be rather straightforward: If trains stopping for these activities could do so clear of the main tracks, capacity would increase greatly. Each solution was designed to accommodate the traffic expected in 30 years, after full development of the program.
The second step involved determining the changes necessary to achieve the goal running time. This part of the program development used the assumption that in terminal areas, passenger trains would operate at conventional speed in order to reduce the need for capacity that is caused by a great differential between passenger train speed and freight train speed. In some cases this involved limiting passenger train speed to less than could be achieved, for example 50 mph instead of 65 mph, because track geometry limited freight train speed to 35 mph. A Train Performance Calculator was used to determine the running time using a maximum speed of 125 mph and also a maximum speed of 110 mph outside of the terminal areas. One line change was assumed. A new route using mostly existing lines was about 8 miles shorter than the current route. Significantly higher speed could be achieved on the new route. The running time difference between 110 mph and 125 mph outside of the terminal areas was just under three minutes. Since the goal transit time was achievable at a maximum speed of 110 mph, the cost of highway grade separations required for 125 mph prompted the decision to make the maximum speed 110 mph.
After the initial running time determination, the route was examined for the track geometry changes required for 110 mph. Realignments and line changes required to support 110 mph that were in environmentally sensitive or difficult to construct areas were eliminated. In doing so, the speed limit in each of these areas was assumed to be 90 mph. 90 mph was chosen to allow freight trains to operate at 60 mph on the entire line. The existing freight train speed limits ranged from 40 mph to 60 mph. Where tracks might be shared because a third track was not necessary merely to support speed, the speed differential would be kept to the same 30 mph (passenger 79 mph freight 50 mph) as the existing situation. Tests with the TPC showed that freight trains could achieve 60 mph over most of the line using the normally assigned power and that a fuel savings resulted from the 60 mph speed limit. The saving was a result of the effect of momentum on ascending grades and the uniform speed limit. After each change, the running time was tested with the Train Performance Calculator. Elimination of 110 mph trackage continued for each difficult location in reverse order of probable magnitude of difficulty as long as the goal transit time could be reached.
Once the location of the necessary 110 mph track was determined, the infrastructure required for passenger train operation was examined. Operation was assumed to be a single track, passenger train only line. Northward and Southward trains at the service goal of 1 hour interval were drawn on separate layers of a stringline. Trains in one direction were left stationary and trains in the opposite direction were time-shifted. The points at which opposing passenger trains met were examined. Meeting points in terminals or other congested areas were avoided to the extent possible. Meeting points where the speed was to be 110 mph were also avoided, since that would require a second Class 6 track. Time-shifting continued until the best set of meeting points were found. At any meeting point, the need for a third main track was assumed. At all points where a third main track ended, crossovers allowing simultaneous movement on any combination of two of the three main tracks were assumed.
Once this entire process was completed, simulation testing of freight and passenger traffic together using DPM was begun. Freight traffic was assumed to be that expected in 30 years, ten years after full program development. From this testing, the need for additional infrastructure to support the freight-passenger combination was determined. After determining the infrastructure required for the entire project, phases in which capital projects could be matched to specific increases in service or reduction in transit time were determined. Each capital project was assumed to be a complete portion of the final project so that no work in any phase would be rendered obsolete and be removed by work in a future phase. A project enabling the addition of only two new schedules would have the ability, in isolation, to support hourly 110 mph service and the freight traffic anticipated in 30 years.
Having a well-developed commercial plan for the passenger service before developing the infrastructure and operating plan allowed the consideration of the most economical and least damaging infrastructure plan. It is understood that freight train operation is not and cannot be as structured as passenger train operation, however it was assumed that freight train operation would be sufficiently structured to avoid the need to queue trains on main tracks outside of yards and terminals.
Copyright 2000 Thomas White
Transportation Planning and Management - Train Dispatching - Home