I never get tired of explaining the simple mathematics that governs safe train operations. Never. Why is that? Well, for one, I'm not very good at complicated mathematics. I'm good at solving for X; I'm good at geometry; I'm good at trigonometry; I'm good with Newton's Laws, and the derivations thereof. Advanced mathematics? Quantum theory? Calculations of objects moving at or close to the speed of light? Not so much.
And for two...?
Well, for two, because the more thorough our grasp of the simple mathematics of time, distance, velocity, acceleration, coefficient of friction, etc., the more sophisticated, deeper, more inclusive becomes our knowledge of what we must do to achieve safe train operations-- the more robust is our calculus of, at, and for the intersection of safety and performance. Good at calculus? Me? Barely. But good at the calculus of safe train operations? Straight As, brothers and sisters, straight As.
So let's get something out of the way right from the getgo. Everything we do, the we being railroad operating officers, is a compromise between safety and operating performance.
We don't operate trains at 100 mph on track that is maintained for 60 mph operation. We don't maintain track at the safety standards set for 100 mph when our operating performance is based on a 60 mph maximum velocity.
My British colleagues would, over a pint or two after work, laugh and say "All signals are red. All trains are stopped. The railway is perfectly safe." Of course, it's no longer functioning as a railway, no longer serving any useful function, but it's perfectly safe.
We calculate risk; we calculate probabilities of events; we reconcile conflicting needs. Based on the determining need and the likelihood of both the occurrence and the severity of an event. We design and deploy systems to manage, reduce, risk while facilitating that determining need. Nothing I ever did was an exception to this calculus.
Grade-crossing protection apparatus is just one of those systems. What is the determining need? To prevent a collision between train and automobile traffic? Actually not. To reduce the likelihood of such a collision is the appropriate, calculated answer. To inform vehicular traffic that the risk of a collison with a train is compounded if drivers stop on tracks or foul tracks. So we identify the crossing with signs. We provide information stenciled on the pavement-- "Do Not Stop on Tracks." At public crossings, we install warning lights supplemented often by gates that descend and act, symbolically, to "stop" a car from fouling the operating area of the railroad.
Of course the lights and the gates can't stop anything really, but that's where the calculus has taken us, given the other determining need.
That determining need is that of the city, county, state, and federal roadway authorities who mandate that train operations will interrupt the flow of vehicular traffic at the crossing for the least duration possible.
We have heard from senators, representatives, experts, civil engineers, that there must be, there are, system that will positively prevent grade crossing collision by automatically stopping a train when the crossing is obstructed.
Amtrak uses quadrant four corner gates and embedded loop circuits on certain crossings. The embedded loop circuits will detect the presence of an obstruction on the crossing. That detection will automatically indicate to, and enforce upon, the locomotive engineer the need to stop the train prior to occupying the crossing.
Does the system work? Sure it does. for Amtrak... for Amtrak running maybe 1 or 2 trains every hour. Sure it does where the vehicular traffic on the road is very light and will not back up, leading drivers to attempt to go around gates and create even greater risks.
But let's think about what happens when we install "automatic stopping" for grade crossings on the Harlem line of Metro-North at Commerce St. in Valhalla.
First off, if we are going to use the signal system to communicate and enforce the zero velocity command, we have to do so in conformity with the design of that system-- that is to say with the required rate of deceleration to conform to the requirement for the zero speed.
Metro-North's signal design utilizes a braking table of standard braking distances for signal spacing. The required rate of deceleration for compliance with signal driven "reduce speed" indications is 1.28 ft/sec/sec. That's a 1.28 ft per sec decline in train velocity every second. Sixty mph is 88 feet per second. The signal design requirement for bringing that train operating at 60 mph to a safe stop means that after 1 second the train is operating at 86.72 ft/sec; after 2 seconds 85.44 ft/sec, and so on.
The distance required to stop a train operating at 60 mph by utilizing the on-board cab signal/automatic speed control apparatus then is 3025 feet plus and additional 704 feet for the on-board apparatus to process the new requirement; for the locomotive engineer to observe and acknowledge the new requirment; for the brakes to be applied to meet the new requirement; and for braking effort of be achieved on every axle of the equipment.
No, railroads do not design their signals, or their signal spacing distance for emergency brake applications, which of course would require less "preview," less time, and distance, to achieve zero velocity. Railroads deploy signal systems precisely to eliminate, or rather reduce the need for emergency braking.
OK, so we set the grade-crossing protection to announce the train when it's more than 3729 feet away, because we want the lights to flash before the gates comes down to give traffic the opportunity to stop. The gates may take 6 seconds to descend, so let's add add another 800 feet to the 3729, so the train gets announced when it is 4529 feet from the crossing.
If the crossing remains clear, wonderful. The train occupies the crossing in 52 seconds. The train is 8 cars in length, and standard length is 85 feet, so the crossing is occupied for a minute.
To release the crossing protection, the entire train has to traverse a release point which adds an additional 2 seconds. The crossing is now released. The gates are lifted, provided another train is not approaching from either direction. The lift itself requires 6 seconds. The railroad has obstructed traffic at this crossing for at least 68 seconds.
Now I just happen to have some old Metro-North Harlem Line employee timetables on hand, from April 2008, and hence woefully out of date as I'm sure the railroad has increased its service to meet the demand of its passengers. But it's the best I can do, and as we used to say back in the day, "Close enough for government work."
So let's see, in the AM for example, between 7 and 830 AM, Metro-North operated 11 southward trains and 4 northward trains.
Given the variables in any operation, we do not presume nor do we require that trains pass each other at Valhalla, so we have 15 trains operating in 90 minutes, each triggering an interruption in Commerce St. traffic of 68 seconds for a total of 17 minutes.
OK, city, county, state, and federal representatives: how prepared are you to have vehicular traffic shutdown for 17 minutes in the morning by train traffic? What do you think the actions will be of drivers stopped directly at the gates when, after 30 seconds, no train occupies the crossing?
Well there are numerous video recordings on youtube and other sites that tell us what the drivers, and pedestrians, do. They run the gates. And run they will.
Now after 30 seconds when the driver decides to run the gate, the cab signal on the approaching train drops from "normal" to "restricting" and because the locomotive engineer is qualified on the territory and the signal apparatus design, he or she knows this indication is the result of an auto obstructing the crossing, so the locomotive engineer places the train into emergency braking because safety is of the utmost importance, and the train only has half the distance it is designed to have for this sort of demand for zero velocity.
Coffees get spilled, and that's messy but as long as nobody gets scalded, no big deal, except for the dry cleaning bills. People get thrown around inside the train, and that can be a big deal which is why we work against emergency brake applications to begin with.
But the train comes to a stop, because even if or after the auto has cleared the crossing, an emergency brake application is irreversible. Now the other drivers at the crossing see that the train is stopped. And guess what they do?
Bulls-eye catnip. They begin to run the gates. Except......except one of those 4 trains coming in the other direction was gradually reducing speed as it was a mile away. Then that second train's cab signals cleared to normal and that second train accelerates, and this time, when cars start to run the gate, this second train cannot stop and guess where we are? We're exactly where we were at Commerce St. on February 3, 2015 at approximately 6:37 PM.
Meaning what works for infrequent train operations does not work for high density commuter service. Meaning that those who think they can enforce zero velocity on trains if the grade crossing is occupied are targeting the wrong vehicle, the wrong mode of transportation. Meaning that those who want to bring the trains to a stop are increasing, not reducing, risk.
Meaning, don't stop the train; stop the cars. It's safer. Embed "tire shredders," like those used by rental car companies in the pavement-- with the apparatus to be bought and maintained by the highway/road authorities. Put your world-famous "inward-outward" video recorders on poles monitoring the crossing, recording license plates of vehicles that violate the crossing warning and suspend their licenses. Interlock all cell phone and GPS functions in autos so that they cannot be accessed by the driver anytime the car is in gear, or velocity is detected. Do what needs to be done and quit trying to gather votes, media time, press clippings, or consulting contracts.
February 7, 2015
"The theory that all negligence which causes serious disaster will always be found to have been more or less habitual is still worthy of respect"--
Railway Gazette, Vol 21, 1889 (with thanks to PN)