Ever since Mr. Westinghouse invented his air brake system, there has been a steady stream of improvements to the design. Here are a few:
- Addition of an emergency braking system consisting of separate valve and reservoir to allowed increased braking force should the train separate accidently or the engineer or conductor in the caboose desire it. If you release air from the trainline at a controlled rate, you get regular braking proportional, service braking. How much de[ends on the amount the pressure dropped. If you release the air from the trainline fast, you get emergency braking. Also, the emergency brake valve will cause the trainline to vent in that car, insuring that the emergency brake application will propagate throughout the entire train. It also adds a layer of safety insuring there is always some braking effort available on the train. Should the engineer unwisely make a service brake application, release it, and then quickly make another application, and repeat this a few times, before the train line and car reservoirs have a chance to recharge, the air brake system could be left with little or no service braking. Emergency braking will still be available.
- Improved brake valve sensitivity. The braking signal gets to each car at nearly the speed of sound. Faster brake valve reaction once the of the apply and release signal from the locomotive is received improved braking response. Also allows finer control of the application and release. An initial application of braking could be followed by a small increase. In passenger service, the release of the brakes can be done gradually, to provide smooth and controlled stops.
-Improved brake valve stability. Internal changes to the valves over the years have resulted in valves that react when they're supposed to and not accidently on their own while allowing train length to grow.
- Many improvements to the design and material of the components to increase reliability, life and reduce leaks. The details are unimportant here, but the cars' brake valves last years between complete functional testing, called a single car test, and rebuild replacement interval is over a decade.
-Empty/load braking. Steel wheels can be counted on to grip steel rails with a force of 20 or 25% of the weight on them without risking sliding under nearly all conditions. A freight car weights about 35 tons and can carry 110 tons of lading. You don't want the wheels to slide when the car is empty, so the maximum braking force is set for that. When the car is loaded, it has much more mass, but no more braking force. A loaded train will take far longer to stop than and empty train for this reason. A valve that senses the spring deflection of the car's suspension can be added that allows increased braking when the car is loaded. This is called a load/empty valve. Currently, it is most often used on unit trains that run entirely loaded in one direction and empty in the other and on intermodal cars where the empty weight is very low compared to the loaded weight.
-The brake control equipment on the locomotive has improved as well. The current state-of the art has replaced a lot of pneumatic valve logic and control with electronics and electrically actuated valves for improved reliability and control. It also more easily interfaces with train signal safety systems, like PTC.
One more thing to add, here. As railroads increased the speed of trains, they needed more braking to be able to stop before the next signal. (there were also changes to signal systems to accommodate speed as well) One of these changes was to increase of the brake pipe pressure, which increases the available braking force - made possible by improved materials that resist leakage. The current standard is 90 psi for freight trains and 110 for passenger trains. Additionally, passenger trains received two additional improvements to allow even faster speed. One was run a wire the length of the train to signal application and release. This is called Electro-Pneumatic Braking (EP). This was rather common for several decades but Amtrak does not use it. The response of standard air brakes is sufficient for the trains, speeds and signal system they operate. The other was a wheel slide detection and correction system similar to anti-lock braking on an automobile that allowed higher braking force without risking wheel slide by measuring the wheel/axle rotational speed and releasing some brake cylinder pressure if the wheel is turning more slowly than its neighbor.
So, that brings us to the latest development. Electronically Controlled Braking (ECP braking). Controlling the brakes on every car electronically by sending data from the locomotive to every car means all the brakes can apply at once, shortening braking distance and improving the control of slack action in the trains. Slack running in and out uncontrolled, under certain circumstances, can lead to or increase the risk of derailments. A data trainline to control the braking can also be used for other advancements unrelated to braking. Tests of ECP braking over the past 30+ years were encouraging.
The problem comes from implementation on mixed commodity trains. If you want to run a train with ECP braking, every car has to be outfitted with ECP braking. A system that fits the existing piping on a freight cars costs roughly $10,000 to buy and install. The fleet of cars in mixed freight service is about 1.2M cars. Total bill = $12B. A lot, but not outrageous.
So a bit about mixed freight service. It runs a lot like UPS handles packages. The cars get sorted out into groups heading the same direction, then combined with others going the same general direction. Along the way, the sorting process is repeated several times until the car is delivered to the consignee. To get an idea of how the actually plays out, on a railroad the size of NS, a carload shipment will move about 500 miles and be carried by five distinct trains. Many shipments move on more than one railroad, so the actual, national average is more than this.
Imagine if half the packages UPS handles are painted blue, and there is a rule that you can't put blue packages in a truck with other packages and you can't put other packages in a truck with blue packages. I either have to add a lot more capacity to each of my sorting facilities so that I have a place for each blue box going to Macon GA and one for each plain box going there or I have to sort only blue boxes today and hold the plain boxes tomorrow. I'd need a bigger fleet of trailers to hold them in until their day to sort.
So, it is with the railroads and sorting of cars for mixed freight trains. If you have to segregate the car fleet into equipped and unequipped, you need some combo of more facilities, track and a less efficient operating plan. All of this is impractical.
Can we take a week and just outfit 1.2M cars? No. First you'd have to produce that many sets of equipment, then you'd have to hire and train quite a few people to do the work, just for one week, then you'd have to find the space to park that much equipment where you could do the work. This is not even close to practical.
Can we outfit each car with both pneumatic and ECP braking and just select each which system is active when we build a train? Sure. But this runs the cost per up many fold as there are brackets and piping and cutover valves that have to be installed on every car. So that $12B might turn into $50B. Can the industry afford that? Maybe. Over many years. It's not a slam dunk. Also, installation pace is slower since the work on each car is a lot more complex. Installation over 5 years means 700 cars a day. Even that's a tough nut to crack.
Four and a half years out, 90% of the fleet has ECP. How often will there be a train of solid ECP equipment? About one in every 70,000 trains will be solid ECP. So, you really have to get all the way to the very high 90s before you can start to run ECP equipped trains.
Are there alternatives? Maybe I run the data connection in every car right away, then add the piping and valves later. I could get to trains with 90% ECP equipped cars sooner, but not much. How many cars without any brakes do I want to allow in a train? The FRA has rules about that....
The inability to handle a segregated fleet and the cost and time lag to handle dual equipment installation has kept railroads from implementing ECP fleet wide.
Let's pivot to cabooses. Remember them? They pretty much disappeared in the 1980s. Their main function was to mark the end of the train and make sure the airbrake trainline was complete and functioning. The person in the caboose had an air gauge connected to the train line and a valve to apply the brakes if something went wrong. It was a terribly dangerous place to ride. If the slack in the train ran in or out, the person in the caboose could go flying. Cabooses were replaced with End of Train devices (EOT) that clamped to the last coupler in the train and had a pressure transducer, a battery and a radio. A receiver in the locomotive showed the engineer what the pressure on the last car was. They also have flashing light, a motion sensor, and more recently a valve to apply the brakes by radio from the locomotive. Many also have an air turbine generator to keep the battery charged. The conductor was sent to ride in the locomotive.
Enough of that.
What next?
My son asks me, "How hard would it be to design a radio controlled emergency air release gizmo that you could insert between every 10 cars? Something that...would be compatible with current technology?"
Light bulb goes off in my head! Not only doable, maybe not that hard at all. And why stop there? Why not do service and emergency braking? Just leverage EOT technology. A radio based - not quite ECP system with available technology!
Here's how. Don't just put an EOT at the end. Put several in the middle of trains. Redesign them to allow for a service rate reduction as well as an emergency brake rate reduction. Reconfigure the data communication scheme to allow more complex commands. It needs to be able to send a command to reduce the air by a certain amount and then let the device use it's own feedback - pressure transducer - to control it.
Some more detail. The device would be installed by the car inspector as he goes down the train connecting the air hoses. The device would mount on the coupler or maybe a bracket on the end one of the cars. The air hose from the device would have a Tee at the bottom and a place to connect each air hose from the adjacent cars to the other "straight" ends of the Tee, so the trainline integrity is maintained. You may also need a small reservoir to reduce transient pressure changes due to air flow - Bernoulli effect if I remember right.
There would need to be a new transceiver on the locomotive to be able to talk with many devices at once instead of just one as is current practice. This transceiver would be integrated with the electronic air brake on the locomotive so that commands to reduce the trainline pressure would come from the locomotive's braking controls.
That's the set up. Let's walk through it.
Car inspector walks the train and installs the devises (MOTs - for Middle of Train). The engineer puts the MOT IDs into the head end box. The trainline is charged to 90 psi, but with normal leakage, the rear of the train is at 80 psi. The pressure at each of the MOTs is somewhere in between. The engineer makes a six psi reduction on his automatic brake valve. This starts venting the trainline at the locomotive as well as sends a signal for a six psi reduction to each MOT. The MOTs react by opening the service rate valve until the pressure is six psi lower than when they started.
This causes a pressure wave to propagate in each direction from the MOT, telling the cars' brake valves to apply the brakes, in normal fashion. Lets say you placed these MOTs every 1000 feet in a 10,000 foot train. Instead of 10 seconds to get the signal to the end of the train, the most the signal has to travel is 500 feet - a half a second. This should provide near-ECP performance in terms of braking distance reduction and slack control.
If you add existing and proven load/empty valves you should be able to reduce stopping distances quite a bit. Perhaps even better than you would get with ECP alone.
Advantages of MOTs versus ECP
No changes to freight cars
Can run with or without MOTs.
Can add more or less to suit
Failsafe. Existing braking system still 100% intact and functional. Since you are still venting the trainline from the locomotives, the brakes will all apply as requested eventually.
Proven technology - EOT radio link. air turbine battery. Existing air brake hardware. All have long term, proven capability
Disadvantages:
Still not "speed of light"
Only braking. Not other advantages of data trainline to each car.
Unknowns:
Radio continuity and data rate between locomotives and cars. Experience with radio signals to mid-train locomotives is a useful guide. Signal can be intermittent. Can be helped with wayside radio repeaters. Would antenna on top or side of each car help? Magnetic mount - use telescoping stick to apply/remove.
Stuck brakes. If you release from the head end before trainline has settled, can you wind up with stuck brakes? It happens now. Would it be worse?
Battery size, hardware and electronics would be more complicated than on EOT.
Could EOTs be modified to try "proof of concept" test.
Is there enough band width in congested areas to accommodate 10x more EOT data traffic on existing EOT frequency?
So, there you have it. Somebody has to have thought of this before and patented it, right? I'm too lazy to look. But, I don't really care. This is an idea worth looking at, I think. This plus empty/load valves on the cars should be able to beat straight ECP in braking distance at speed. Probably, less effective at very low speeds since brake valves are still all pneumatic control.
I would think it would only take a few months to modify some EOT equipment to do a proof of concept test. If successful, maybe a few years to become an industry standard product.
What do you think?
