Capacity is where you find it: how BNSF balances infrastructure and operations

Railway Age, Feb, 2005 by William C. Vantuono

Burlington Northern and Santa Fe's 2,217-mile Chicago-Los Angeles Transcon is one of the busiest, most fluid rail routes in North America. Freight trains operating at throughput levels approaching 100 per day on some portions, and at speeds as high as 70 mph (for premium intermodal services) roll on a railroad that's 93% double-track and 100% CTC equipped. Ideal infrastructure conditions and plenty of line capacity, one might think?

Not exactly. The 3% (63 miles) of the Transcon that's getting double-track this year, and the remaining 4% (86 miles) slated for 2006 and beyond can, under certain conditions, become a bottleneck that affects the entire operation. But single-track segments aren't the only limiting factors that BNSF considers as it deals with the daily challenge of capacity and fluidity on a network where traffic has been on the rise for the better part of two years.

There are no fewer than ten factors that impact line capacity, says BNSF Vice President-Service Design and Performance Rollin Bredenberg, who leads a team with one of the most challenging, rewarding--and perhaps frustrating--tasks in freight railroading: getting business that's growing and changing to function on infrastructure that's fixed--and expensive to expand or modify. The big ten, which combined can be as complex as a scientific formula or mathematical equation, are these:

* Number and length of trains.

* Horsepower of trains.

* Peak train counts, delineated by day of week and time of day.

* Terrain (grades and curvature).

* Variance in train speeds and dispatching priorities.

* Track speeds (which can change from day to day, depending upon maintenance requirements). Types of signal systems.

* Amount of double- and triple-track mileage.

* Length and spacing of sidings.

* Terminal infrastructure.

In the case of infrastructure, is more of it better? Bredenberg's rule of thumb: "Build it as a last resort." This is the concept an industry struggling to earn its cost of capital has been more or less forced to embrace. In some cases, as on the Transcon, getting to an ideal infrastructure level--100% double-track, 100% CTC, double crossovers every, 12-14 miles on subdivisions where traffic is heaviest (the Oklahoma/Texas Panhandle Subdivision, for example), even some triple-track (Cajon Pass)--takes years of work, limited by available capital. Those last 86 miles of double track in New Mexico, the Panhandle Subdivision, and Missouri, for example (map, above), are described by Bredenberg as "high unit cost," as they will involve construction in difficult terrain.

Service Design has a variety of methods and tools at its disposal for capacity planning. They include:

* Historical trend analysis incorporating run time, train length, and horsepower per ton calculations.

* Dispatch system replays (made possible by DigiCon's Digital Traffic Control system installed at BNSF's Fort Worth, Tex., Network Operations Center).

* Marketing forecast information.

* Identifying available alternative train routings.

* TPCs (Train Performance Calculators), incorporating such elements as locomotive fuel burns and unimpeded train running times.

* Train dispatching simulators. Much of the Service Design group's analytical work is performed using RTC (Rail Traffic Controller dispatch simulation software and models from Berkeley Simulations.

Then there are such inevitable operational "events" as train bunching, service delays, and grade crossing collisions. In a perfect world, says Bredenberg, train movements are spaced as evenly as possible and stay that way. But that rarely happens.

At what point does it become necessary to add infrastructure, and at what levels? Bredenberg describes five network capacity "thresholds," beginning at 60 trains per day average:

At an average of 60 trains per day and a peak of 75, double-track must be added to single-track segments where there are already 8-10 miles between sidings. Also, 60 trains per day is the absolute capacity limit for "CTC islands" (CTC-controlled bi-directional segments at interlockings separated by unidirectional automatic block signaling), says Bredenberg. Above 60, "there is no substitute for CTC."

Bumping this up to 70 trains per day, and a peak of 85, on existing double track requires universal (double) crossover plants every 10-12 miles, using No. 25 (40 mph) turnouts. This is how trackage on the Needles, Seligman, and Gallup subdivisions has been configured. At 85 average trains per day, and a peak of 100, add double track where there are 2-8 miles between sidings or double-track sections. These two thresholds are typical of the Transcon.

When traffic averages 100 trains per day and peaks at 115 on existing double track, triple-tracking is required for segments with very steep grades, slow track speeds, and/or heavy passenger train traffic, in order to enable overtakes. The best example of this is Cajon Pass, which is shared with Union Pacific.

Once traffic reaches an average of 120 trains per day and peaks at 135 on existing double track, triple-tracking is required on additional heavy grade locations. Such infrastructure is employed between Barstow, Calif., and Clovis, N.Mex. At this density, adds Bredenberg, it's important to avoid queuing trains outside of terminals.