How CN does it That is, run a scheduled railroad. And why? As Hunter Harrison points out, "If we provide better service, we'll get paid for it." By William C. Vantuono, Editor Photo by Canadian National For most concepts in business or in life, either you get it, or you don't. When it comes to providing good service, some railroads just don't get it. Some railroads do. Canadian National gets it, and makes money at it, too. How? Two words that, up until recently, too many railroaders thought created an oxymoron: "scheduled railroad." Running a scheduled railroad is not a new concept. It's not easy to do, but neither is it rocket science. The railroader who's made it at least look easy is CN Executive Vice President and Chief Operating Officer Hunter Harrison, who runs what is arguably North America's best-performing Class I. CN has the numbers to bear this out: ontime performance consistently better than 90%, an operating ratio below 70%, and plenty of satisfied customers. This didn't happen overnight. "Most of us who have spent a lot of time in this industry were not taught scheduled railroading concepts," says Harrison. "We were taught the old, conventional wisdom: When you get a big train together, run it. In a regulated industry, everything is cost-driven. We talked about service but didn't do anything about it. Moving to a concept that said good service is going to drive costs down and improve asset utilization caused people who for years had been taught otherwise to start shaking their heads. It was a difficult task to get the message out and have people understand it." To some degree, CN is still going through this "cultural change," as Harrison describes it. "As much as I would like to brag-and we've done very well-have we gotten to the point where we can't bring this to another level? Absolutely not. We have a long way to go." What is the next level for scheduled service? "To continue to improve what we do," says Harrison. "To improve velocity and become more consistent. It's the right thing from the customer standpoint, from a cost-control standpoint, from an asset utilization standpoint. It's the right thing for our employees as far as quality of life is concerned." Operating employees have "a more predictable, stable work environment," owing to the fact that trains generally operate to plan, regardless of traffic levels. This protects power and crew cycling. Train crews know far in advance what trains will run. "Balancing" trains by direction reduces deadheading, enabling crews to gain more miles by working in both directions on most trips. Operating costs associated with unproductive time are reduced. Hunter Harrison's seven steps to superior service-and a better bottom line At Canadian National, scheduled railroading is based upon "a few simple principles," according to Hunter Harrison, who outlined them in a recent Railway Age Point of View column (December 2000, p. 14). They're worth summarizing: 1 Minimize car dwell time in yards. CN districts have standards for connections and maximum terminal throughput. These standards, along with switching windows at origin/destination terminals, anchor main line train arrival and departure times. By bringing the connection time for major traffic flows as close as possible to the design standard, transit time, car cycles, and yard congestion are reduced. 2 Minimize classifications. The best way to reduce idle time in intermediate yards is to minimize the number of times a car is handled. CN pays particular attention to the number of times cars are classified, with an objective to have no more than two classifications per car. 3 Use multiple traffic outlets between yards to keep traffic moving. CN does this by minimizing the number of blocks created at origin, and moving the traffic on a conveyor-like service to major classification yards or hubs. This provides flexibility to reduce or add capacity by changing the frequency of trains on the satellite/hub conveyor. 4 Run general-purpose trains. Some cars that would normally be handled in a unit train are placed in general-purpose trains to fill idle capacity. Unit trains exceeding 48 hours to load or unload or less than maximum train length are evaluated for conversion to general-purpose freight. Also, combining intermodal and carload trains drives up trainloads and creates the opportunity for destination trains. 5 Balance train movements by direction to reduce power and crew deadheading. To reduce imbalances caused by day-of-week volume fluctuations, CN trains are designed to run long (over-siding in the high-demand, or "head haul", direction). In addition, general-purpose trains are used to ensure maximum loading of trains moving in the head-haul direction. 6 Minimize power requirements. CN balances train velocity against horsepower-per-ton (HPT) requirements. Where small reductions in HPT reduce the number of locomotives needed, and the velocity impact does not negatively affect service, HPT is reduced. Since a balance has to be maintained, locomotive reductions are only considered for head-haul trains. 7 Space trains to support a steady workload flow through various yard processes. This improves yard efficiency, reduces the number of assets required, and decreases the cascading effect of failures. CN's scheduled railroad is based upon seven basic concepts (sidebar on right). Right now, only carload and intermodal services are fully scheduled. But unit trains (coal, grain) "are dealt with philosophically in the same way, as far as asset utilization and controlled cycles are concerned," says Harrison. This has resulted in a departure from "standard" Class I operating practice. For example, says Harrison, "We used to run a high percentage of our grain business in unit trains. We'd sit and wait-maybe four or five days-to accumulate 100 or more cars for a unit train, because the conventional wisdom said a unit train is a low-cost operation. Even though the rest of the railroad was 'balanced,' we were still deadheading crews and power." Harrison says he and some of his people realized there was a better way to run a railroad. "We saw we were loading 15 to 20 cars of grain per day in some markets. We said, why wait and accumulate them, when our existing merchandise trains-which have a schedule-are running at 70% to 75% capacity? Let's take those grain cars and put them on a merchandise train. It fills that train out to capacity, lowers the cost of the grain and the merchandise, improves asset utilization, and reduces deadheading." Today, as much as 50% of CN's grain business does not move in dedicated unit trains. "That flew in the face of conventional wisdom," says Harrison. "But some people who used to feel strongly about conventional wisdom don't feel so strongly anymore." Harrison is quick to point out that he is not criticizing all unit train operations. But there's a difference between what he calls "true" unit trains and trains that sit around collecting cars until the railroad thinks they're big enough to run. "A true unit train is a wonderful operation, like Powder River Basin coal trains," he says. "You take 110 cars in there, load them, go to the utility, dump them, and go do it again in a fast, closed loop-that's wonderful. But when you create a grain train through holding, and you go to the port and the cars unload 15 to 25 at a time, and then the train comes apart as the cars are redistributed-that's not a unit train." The philosophy of operating a scheduled railroad has produced a ripple effect throughout the railroad. For example, it has allowed for better scheduling of CN's fast intermodal shuttles. But, "it's not just scheduling trains and cars," says Harrison. "It's a whole operating philosophy of balancing assets and costs. It involves scheduling all the processes that lead up to the movement of trains and cars. If trains are running on schedule, you can predict where they're going to meet. You can look at line capacity issues. It's far more involved than a lot of people give it credit for." CN, says Harrison, has not relied heavily on computer modeling to implement a scheduled railroad. "It's more philosophy, common sense, and good judgment," he says. "We did use some tools that helped us make the process less cumbersome than it can be, but do we have a lot of models with all these algorithms that tell us what to do? No. We haven't gotten that far yet. For the people involved, it's a tremendous learning and development process. So even if a model could do it for us, I'd still want to do most of it manually, for the value it gives to the people developing the schedules." What about interchange traffic, particularly scheduled CN trains that connect with less-than-reliable U.S. Class I's? "It's a problem, and we don't have it solved," says Harrison. "The first thing we tell a customer is that we can schedule a car on our railroad, and we hope our partner does what he says he's going to do. We work individually with certain railroads to schedule certain movements." Harrison says CN is actively engaged in spreading the gospel of scheduled railroading. "We have no pride of authorship, but we think we have something here," he says. "We're trying to influence our railroad partners to work with us. If we can all schedule, cut down congestion at gateways, we'll gain market share. We'll all benefit." Norfolk Southern recently announced it's going to reconfigure its network, which has been limping along since the Conrail acquisition, to a scheduled operation. It's widely believed that it will be closely patterned after CN. As for some of the other Class I's, CSX Transportation is leaning in that direction, and Burlington Northern and Santa Fe is pursuing it in the merchandise area, with measurable success. "There's a shift occurring in the industry," says Harrison. "We all did this, more or less, with intermodal. We all do a pretty good job running on line-of-road, from point A to point B-that's not too difficult. It's terminals-dwell time, matching up, scheduling-that gets difficult. That's where the opportunities are. When people were talking about scheduled railroading, not everyone liked it. But when somebody did it, and had the lowest operating ratio, heads started turning. It's amazing how cost still drives this industry, to a degree." That "somebody" was the Harrison-run Illinois Central, which was merged meltdown-free into CN in 1999. At the time, IC had the lowest operating ratio among U.S. Class I's. Harrison, whose operating expertise was immediately tapped by CN President and CEO Paul Tellier, extended the scheduled railroad concept to the entire network. Critics said it was an easy thing to do, pointing out that CN and IC were simple operations, compared to most of the other large carriers. "I don't hear that anymore," says Harrison. Harrison's concept of scheduled railroading dates back to his days in the late 1980s as the Burlington Northern's regional vice president in Chicago (former CB&Q territory), an area "which is about as complex a network as you can get," he says. "[former BN President and CEO] Darius Gaskins was the driving force behind this, and he coined the term 'service by design.' It basically meant, if we provide better service, we'll get paid for it. The issue was, could we control costs and improve asset utilization at the same time? When we did, it was really gangbusters, but it took a while for people to believe it." That philosophy "died in the early '90s, then was resurrected at BNSF in the late '90s," says Harrison. Before their aborted (or perhaps postponed) merger, CN and BNSF "spent some time on this, and BNSF understood the concept. The merger didn't go through, but it opened our eyes to some things that we should have been doing-whether we merged or not." Wisconsin Central, which pending STB approval will become part of CN, "is similar to us, very service-sensitive. They see the benefits of a scheduled railroad, and hopefully it will kick in within a few months on their system." The greatest benefit of running a scheduled railroad, says Harrison, is that "as you improve service and equipment velocity, it lowers its cost. Revenue-wise, anything that was marginal becomes less so." The end result, he says, is that a railroad can start selling on service-not price. "It's taken time for CN to develop credibility in the marketplace, but we're starting to get rewarded," he says. "We'll continue to raise the bar. This country is begging for rail service to move up to the next level." -------------------------------------------------------------------------------- Copyright © 2001. Simmons-Boardman Publishing Corp.
If there is any RR in trouble it's the Esquimalt & Nanaimo Railway. With the current amount of traffic on the line there's no way it'll survive unless Rail America does something quick.
