Wednesday, April 28, 2010

PTC: Is everyone on board?

Written by  William C. Vantuono, Editor-in-Chief
The Class I’s are, and the technical cooperation, says one key participant, “is truly remarkable.”
Throughout the past 25 or so years, as the railroad industry experimented with various iterations of advanced train control, the most problematic issues have revolved around technology, specifically, its feasibility: Vital (safety-critical) or non-vital? Wayside-, onboard-, or office-centric? How to provide accurate train location and positioning? What type of communications backbone? What type of cab display? How to establish interoperability standards? And—most problematic—what’s PTC going to cost, and is it worth the price?

The Rail Safety Improvement Act of 2008, which mandated system-wide Positive Train Control by Dec. 31, 2015 on all lines where passenger and/or TIH (toxic inhalation hazard) traffic operates, prompted the industry to consolidate what had been a series of disjointed and often proprietary projects into a coordinated effort. As work progresses on developing interoperable technology, the public face of PTC has become the debate on its cost-benefit ratio and how much capital would have to be diverted from other areas to pay for it.

What do cost-benefit studies show? A cost-benefit ratio of 20:1 has been widely used, though few seem to understand precisely what that means. The latest intelligence on this front came late in April from a study by Oliver Wyman commissioned by the Association of American Railroads. As the AAR said in a statement on April 27: “The Oliver Wyman study concluded that the upper limit of business benefits from PTC is $413.2 million over 20 years, using a 7% discount rate. The Federal Railroad Administration assessed the potential safety benefits of PTC as $440 million over 20 years, using a 7% discount rate. Taking the two into account, the maximum safety and business benefits from the PTC mandate is $853 million over 20 years. Given that the FRA has assessed the cost of PTC installation to be $9.55 billion over 20 years, the best-case scenario would put the cost-benefit ratio at 11 to 1.”

So in effect, the industry is united in saying, “Safety is our first priority. Yes, we’ll get it done—but understand that PTC is going to negatively impact us in other ways.” This may be particularly true for Class II and III railroads.

The “get it done” part has been quietly going on behind the scenes, carried out by people with a mission as well as a mandate. “There are so many moving parts,” says Union Pacific Assistant Vice President Transportation Systems Jeff Young. “PTC is well beyond the complexity of anything we’ve ever done as an industry. The deadline is Dec. 31, 2015, but what many people don’t realize is that, in order to meet the deadline, we have to have our systems ready for testing and certification by late 2012. There’s a very high risk involved in that.” PTC implementation plans were filed with the Federal Railroad Administration on April 16, and the FRA has 90 days (by July 16) to respond to them. Now, as they say, the fun begins.

The risk and complexity that Young refers to lies not so much with the individual PTC platforms themselves—Vital Train Management System (UP), Optimized Train Control (Norfolk Southern), Communications-Based Train Management (CSX), Electronic Train Management System (BNSF)—but with interoperability. “We’re confident that VTMS is mature,” says Young of UP’s system. “Interoperability—the ability to assure handoffs from one railroad to another at operating speed—is another story. It’s not going to be easy, and many people don’t understand the complexity.”

To tackle interoperability, the “Big Four”—UP, NS, CSX, and BNSF—have come together under the auspices of the Association of American Railroads and formed the ITC (Interoperable Train Control) Working Committee. Young, chairman of the AAR PTC Policy Committee, says “the consensus has been truly remarkable.” ITC involves no fewer than seven closely coordinated Technical Teams: Applications, Architecture, Communications, Data Management, Messaging, On Board Locomotive, and Wayside. ITC began as the result of an agreement, signed by the Big Four, to establish PTC interoperability standards. The scope of work includes developing uniform interface standards, standardized messaging formats, wireless protocols, braking algorithms with interoperable hardware platforms, wayside and base station equipment, and a locomotive platform that will enable infrastructure sharing (run-through operations). All of this is based on utilization of the 220 MHz frequency spectrum. Canadian Pacific, CN, Amtrak, Kansas City Southern, and other railroads that contribute what the ITC Working Committee refers to as “qualified technical resources” can participate on the ITC Technical Teams.

The ITC rollout schedule is rather ambitious, as the Working Committee’s plan is to have a system revenue-service ready by Jan. 1, 2011. The ITC protocol is currently in the specification phase and will soon be in testing. Following that, WIU (Wayside Interface Unit) hardware will be available for installation. Next comes testing of messaging and communications infrastructure. By late 2010, production radios are expected to be available, with all WIU/radio updates completed. To date, the Working Committee has reviewed more than 250 functional issues that have accumulated since late 2007.

Key to interoperability is commonality. Such elements as a system requirements specification, locomotive-to-office and locomotive-to-wayside ICDs (Interface Control Documents), single onboard software-executable common consist data, and a track database file must be common to all PTC systems. The scope of work for PTC radios that support interoperable train control systems and functions involves development of 220 MHz radios for locomotives, wayside equipment, and base stations; design of hardware and “air interfaces”; development of physical and “link layer” protocol software; and implementation of components that will run on 220 MHz radio hardware.

Given all the uniform standards and architectures being developed, it’s reasonable to assume that, in addition to improving safety, PTC will improve operations and free up capacity, right?

Wrong, say the experts. PTC could—again for complex reasons not fully understood by many—actually decrease capacity. (The FRA does realize this, and included a statement to that effect in its 2009 PTC Notice of Proposed Rulemaking.) This has mainly to do with the advanced microprocessor-based speed control and braking algorithms that provide the foundation for PTC—and how they override the human element of a skilled locomotive engineer at the controls. Unlike a locomotive engineer, PTC systems as currently configured cannot deal with ambiguity, so a safety buffer must be built into the system. The size of that buffer, which is still being determined, could be far bigger than needed. Add to that variations in weather conditions and train handling, and what you’ve potentially got is an overly conservative technology that automatically assumes a worst-case operating scenario and thus could very well slow a train below the level at which it can be safely operated by a skilled engineer. Railroad operating speeds are typically calculated with a safety margin so that minor excesses in speed will not result in an accident. A PTC system with a more conservative safety margin could decrease performance by contributing to a drop in the hard-fought-for number known as system velocity. (TTCI is currently working on methods to enable braking algorithms to adapt to variable operating conditions.)

Then, of course, there’s the amount of onboard hardware that’s required—train management computers, display screens (FRA requires two in a cab), interfaces to OEM control equipment, wireless internet, GPS gear, etc. etc. “Any one of these devices breaking down forces PTC into failure mode,” says UP’s Young. “Absolute block protection takes over. This imposes a maximum operating speed of 25 mph in dark territory and 49 mph in signaled territory. What this means is that unless a PTC system is 100% reliable, it will impact operations.”

The only type of PTC system that could increase capacity is one that incorporates dynamic moving block, also known as virtual block or flexible block. In simple terms, this is a safety “envelope” that moves with the train. The FRA rulemaking allows for three types of PTC: a vital overlay (the UP, CSX, and NS systems); a non-vital overlay (the BNSF system); or stand-alone moving-block. The last “is the next development in PTC, after the industry has met the 2015 mandate,” says Young. PTC systems that incorporate moving block have been devised, he points out, “but they have typically been over-engineered and far too expensive. Also, migrating from an overlay to a stand-alone moving-block system would require FRA recertification.”

PTC does have the potential to become more than a mostly unfunded safety mandate, and move beyond what the law requires—prevention of train-to-train collisions, overspeed protection, enforcement of civil speed restrictions, detection of misaligned turnouts in dark territory, and roadway worker protection. In its most sophisticated form, it could enable many levels of automation—paperless track warrants and work orders, for example—and become the backbone of a truly “intelligent” rail network. But before the industry gets to that level, it has to figure out how it will achieve, and pay for, the basics.