Super Sidings

[n012944]

     If I'm doing the math right, a train moving 40 mph going into a 20,000 foot siding only takes 5 minutes and 40 seconds until it pops out the other end.  Take into account, that a 6,000 train would take 1 minute and 42 seconds to completely leave the main, and you seem to have only about  4 minute window, to run a train through a slot.  Since that train will take approximately 1 minute and 42 seconds to pass the spot as well, it seems you are working on some pretty precise timing-no?

You're on the right track (bad pun).

Conversion factor for mph to feet per second is 1.467, so at 40 mph, train is going 58.68 fps so yes, 5.7 minutes.

Braking distance 40 to 0 for 0.0 gradient is 5,488 feet with 143 tons per operative brake.  So if we're to avoid braking the train, we have only 14,512 feet to work with.  But the train entering the siding, assuming it has a mid-siding signal, can't get better than a R/FY aspect (depending on the railroad's particular signal rules) because the signal at the other end is red.   

A 6,000 foot train actually needs to travel at least 6,500 feet to get past the insulated joints  and clear the O.S. for the turnout, or 1.85 minutes.  Add .33 minutes for the switch machine to move from reverse to normal (let's assume the DS has the route stacked).

You can see where this is going. 

RWM 

Railway:

Stop with the calculations. Tell me how many times does a train on a double track siding with 6,000 feet or 10,000 feet has the chance to meet at 45mph.   It doesn't take a Ph.d to figure this out.

6000ft, Never  10000ft once in a while,  20000ft, as discussed in the article, several times a day.

[chefjavier]

     If I'm doing the math right, a train moving 40 mph going into a 20,000 foot siding only takes 5 minutes and 40 seconds until it pops out the other end.  Take into account, that a 6,000 train would take 1 minute and 42 seconds to completely leave the main, and you seem to have only about  4 minute window, to run a train through a slot.  Since that train will take approximately 1 minute and 42 seconds to pass the spot as well, it seems you are working on some pretty precise timing-no?

You're on the right track (bad pun).

Conversion factor for mph to feet per second is 1.467, so at 40 mph, train is going 58.68 fps so yes, 5.7 minutes.

Braking distance 40 to 0 for 0.0 gradient is 5,488 feet with 143 tons per operative brake.  So if we're to avoid braking the train, we have only 14,512 feet to work with.  But the train entering the siding, assuming it has a mid-siding signal, can't get better than a R/FY aspect (depending on the railroad's particular signal rules) because the signal at the other end is red.   

A 6,000 foot train actually needs to travel at least 6,500 feet to get past the insulated joints  and clear the O.S. for the turnout, or 1.85 minutes.  Add .33 minutes for the switch machine to move from reverse to normal (let's assume the DS has the route stacked).

You can see where this is going. 

RWM 

Railway:

Stop with the calculations. Tell me how many times does a train on a double track siding with 6,000 feet or 10,000 feet has the chance to meet at 45mph.   It doesn't take a Ph.d to figure this out.

[Anonymous]

In my experience they like to use sidings like that for staging trains. Coming into Vancouver (WA) from Pasco the last 2 or 3 miles (cant remember exactly) double track. Allows a closer place for trains to wait to get into the terminal. They tie down a lot of d.o.l. trains that there too.

[anb740]

     anb740:  Thanks for the input.  Any idea what kind of speeds the trains would be doing during these operations?  Thanks

 On the double track/super siding segments, all of the switches are good for 45mph....and the engineers will do every bit of it.  If you only have two opposing trains meeting each other, the crews and dispatcher can time everything for the perfect running meet.  This actually happens a lot more than you think.  However if you throw more trains into the mix like this morning, someone will eventually have to stop.  But the wait is usually brief when dealing with a seasoned dispatcher, and the stopped trains are cleared for 45mph back onto the single main after the "opposition" goes by....that's whole lot better than inching out of a 15mph siding!

[Railway Man]

Ok, here's a real example for you on the advantages of super-sidings/double track.  The image below is an ATCS screenshot of the same line discussed in the Eastern Corridor article (CSX Fitzgerald Subdivision) taken this morning. 

http://www.georgiarailfan.net/images/atcs1-4.JPG

At the time, there were nine trains between Cordele and Manchester, GA; that's actually a pretty small number given how many usually run during the day.  Note the double track between Mauk and Rupert, GA.  Two northbounds were stopped on track 2 waiting on two southbounds to pass. The first one (N160) was later crossed over at Dyson onto track 2 to allow trailing Q549 to go around him. (the current train meet record for this stretch of double track was 7 trains between the switches)  Had the original 8,800' siding still been in place at Rupert with 15 miles of single track in between, all of this would have been impossible.  Not to mention the huge time delays being racked up because the northbounds had to stop further back in other sidings for the meets.  So, these super sidings aren't just good for speed; they're also good for cramming more trains between two locations with greater ease as Joe Mc pointed out.

Now put yourself in the shoes of the accountants.  Would you rather have a train go through 2-3 crews leapfrogging from one 15mph siding to the next for 200 miles, or would it be more cost effective to do running meets for most of the way on strategically placed double track.  Run the numbers on that one....it took CSX several years to figure it out, and they finally wised up and did something about it!

A fascinating look at the railroad at that moment in time, and an operating/investment strategy worthy of study to see if it is better than alternatives on this particular line segment, and under what conditions it is better.  Thanks for posting this thread and this image!

RWM 

[Murphy Siding]

     anb740:  Thanks for the input.  Any idea what kind of speeds the trains would be doing during these operations?  Thanks

[anb740]

Ok, here's a real example for you on the advantages of super-sidings/double track.  The image below is an ATCS screenshot of the same line discussed in the Eastern Corridor article (CSX Fitzgerald Subdivision) taken this morning. 

http://www.georgiarailfan.net/images/atcs1-4.JPG

At the time, there were nine trains between Cordele and Manchester, GA; that's actually a pretty small number given how many usually run during the day.  Note the double track between Mauk and Rupert, GA.  Two northbounds were stopped on track 2 waiting on two southbounds to pass. The first one (N160) was later crossed over at Dyson onto track 2 to allow trailing Q549 to go around him. (the current train meet record for this stretch of double track was 7 trains between the switches)  Had the original 8,800' siding still been in place at Rupert with 15 miles of single track in between, all of this would have been impossible.  Not to mention the huge time delays being racked up because the northbounds had to stop further back in other sidings for the meets.  So, these super sidings aren't just good for speed; they're also good for cramming more trains between two locations with greater ease as Joe Mc pointed out.

Now put yourself in the shoes of the accountants.  Would you rather have a train go through 2-3 crews leapfrogging from one 15mph siding to the next for 200 miles, or would it be more cost effective to do running meets for most of the way on strategically placed double track.  Run the numbers on that one....it took CSX several years to figure it out, and they finally wised up and did something about it!

[joemcspadden]

One of the chief, if more mundane, advantags of super sidings is the
fact that 4-train (and even sometimes 5-train) meets can be staged.
When things start going wrong on a busy single-track main, this
can be a very valuable relief valve.

Joe Mc

[Railway Man]

It can happen.

Try this for a scenario.  We'll use a four-aspect route-signaled railroad because the aspects are much simpler than speed signaling.  Signal aspects and indications are:

1. Red (indicates stop)

2. Yellow (reduce speed to 30 mph, prepared to stop at next signal)

3. Flashing Yellow (reduce speed to 50 mph, prepared to pass next signal not exceeding 30 mph)

4. Green (maximum authorized track speed)

Entering aspects to a siding are R/G, R/FY, and R/Y depending on the condition of the switch at the other end -- if the switch at the other end is normal the best possible aspect on entering is R/Y; if it's reversed and there's nothing in the next two blocks the best aspect on entering is R/G. 

Let's say our railroad is perfectly flat so that our braking curves are identical in either direction, we have a 20,000' siding (between insulated joints that has #24 movable point frog turnouts that are good for 50 mph, and our two trains that are going to meet here are 143 tons per operative brake and limited by rule to 50 mph.  Both are 6,000 feet long including locomotives.  It's essential that we install a mid-siding signal to break the siding into two 10,000 blocks, because it will give us much more favorable aspects to help make our non-stop, non-braking meet come true. 

The best possible scenario for a non-stop meet is that each train arrives at the last intermediate signal before the siding at the same time and both are running at 50 mph.  For the train that will hold the main, his aspect progression at that moment is G, FY, Y, R.  The train that will take the siding has G, R/FY, Y, R.  Neither engineer makes a control change.

Now the trains approach the absolute signals at end of siding and one train sees FY and the other R/FY.  Neither has to reduce speed because both are already at 50 mph and the turnout is good for 50 mph.

Next, each train sees a hard yellow and must reduce speed to 30 mph.  Let's assume sight distances are commensurate with our 50 mph speeds so each train gets a 4,000 foot look at the signal before he reaches it.  Each train has just left the O.S. behind it (because they're 6,000 feet long in a 10,000 foot block).  In about 10 seconds the switch that's reversed will become normal, and the switch that is normal will reverse, and the yellow signals each engineer sees will go to green.  Neither train initiates braking and the two pass at 50 mph.

That's perfect world.  Real world isn't flat and straight and two trains aren't likely to space perfectly, and not too many engineers with a 143 TOPB, 6,000 foot train, are going to look at a hard yellow and not go into dynamic or make a reduction.  The safe stopping distance with 143 TOPB at 50 mph with 0.0% grade is 7,600 feet.  The real train will usually stop shorter than that but you don't locate signals for the condition you hope for, you locate for the condition you might have. 

Real world one or both trains has to brake down to 10-20 mph.  Still better than 0 mph!

Critical to this scenario is the mid-siding signal and the #24 turnouts.  If we have #15 turnouts good for only 30 mph, the train that holds the main will almost always get knocked down to 30 mph as well.  Of course if our trains are shorter and lighter -- say 50 TOPB -- then we have a better chance of either one not braking.

Is the 20,000 foot "super siding" spaced every 20 miles better than the 10,000 "plain old siding" placed every 10 miles?  It utterly depends on your railroad's alignment and traffic, and what you want your railroad to do.

RWM 

[jeaton]

Quite clear that it would be great timing or just plain luck to have opposing trains meet and pass at the siding with both being able to maintain 40MPH or higher speeds.  On the other hand, if the first to arrive took the siding and after clearing the switch was down to a speed of 10MPH, it would be another 15 minutes before the engine was at the signal at the far end of the siding.  With that, the chance of having the first train to arrive at the siding come to a stop is reduced.  On the other hand, one train having to stop on an 8-10,000 siding seems almost inevitable.  Is that much of an added benefit to the super siding?

[Railway Man]

     If I'm doing the math right, a train moving 40 mph going into a 20,000 foot siding only takes 5 minutes and 40 seconds until it pops out the other end.  Take into account, that a 6,000 train would take 1 minute and 42 seconds to completely leave the main, and you seem to have only about  4 minute window, to run a train through a slot.  Since that train will take approximately 1 minute and 42 seconds to pass the spot as well, it seems you are working on some pretty precise timing-no?

You're on the right track (bad pun).

Conversion factor for mph to feet per second is 1.467, so at 40 mph, train is going 58.68 fps so yes, 5.7 minutes.

Braking distance 40 to 0 for 0.0 gradient is 5,488 feet with 143 tons per operative brake.  So if we're to avoid braking the train, we have only 14,512 feet to work with.  But the train entering the siding, assuming it has a mid-siding signal, can't get better than a R/FY aspect (depending on the railroad's particular signal rules) because the signal at the other end is red.   

A 6,000 foot train actually needs to travel at least 6,500 feet to get past the insulated joints  and clear the O.S. for the turnout, or 1.85 minutes.  Add .33 minutes for the switch machine to move from reverse to normal (let's assume the DS has the route stacked).

You can see where this is going. 

RWM 

[Murphy Siding]

     If I'm doing the math right, a train moving 40 mph going into a 20,000 foot siding only takes 5 minutes and 40 seconds until it pops out the other end.  Take into account, that a 6,000 train would take 1 minute and 42 seconds to completely leave the main, and you seem to have only about  4 minute window, to run a train through a slot.  Since that train will take approximately 1 minute and 42 seconds to pass the spot as well, it seems you are working on some pretty precise timing-no?

[mudchicken]

Not really.  With the longer siding (with 50mph turnouts), trains are able to meet opposing trains and keep up their speed without stopping.  And when you're trying to cram 40+ trains a day onto 200+ miles of trackage, timing is a big issue.  Case in point....here are two trains passing each other at 45mph on the new "Super Siding" at Lilly, GA. 

http://anb740.rrpicturearchives.net/showPicture.aspx?id=507198

Had this meet occurred on the original 6600' siding here, the northbound would have had to stop and eat a minimum 45 minute delay.  And three years ago before this expansion took place, crews were guranteed 12 hour workdays with up to two recrews just to get one train from Waycross to Manchester!

at 45mph, you run out of real estate real quick at 10000 feet. Better hope that opposing train isn't dropping signals in your face and is still close enough to make a clean running meet. If you get too many trains out there, you're buried in the hole.

[BaltACD]

 you wish you'd taken up a more relaxing job such as alligator taunting. 

RWM

Or sales!

     Wouldn't this take a fair amount of luck, to time a moving pass ( or whatever it's called ) or meet, on a relatively short double track, at 40 mph?

For the most part, Dispatchers communicate with trains about what trains they will be meeting and where they will be meeting them.  Crews can tell you to the minute when they will get to Point Q from where they are now with the train they have.  When the Dispatcher tells train 1 and train 2 that the are going to meet at Q both trains know how long it will take them to get to Q, generally at some point prior to the meet Trains 1 & 2 will contact each other to find out their exact locations and the operate their trains so they can have a rolling meet.  Happens every day.

[Murphy Siding]

 you wish you'd taken up a more relaxing job such as alligator taunting. 

RWM

Or sales!

     Wouldn't this take a fair amount of luck, to time a moving pass ( or whatever it's called ) or meet, on a relatively short double track, at 40 mph?

[Railway Man]

     The article Eastern Corridors, in the Feb. 2008 issue of Trains Magazine notes a section of CSX track built with 'super sidings'.  They are 20,000 long, double track, with multiple crossovers.  Why would these be any better than have 2 or 3(?) shorter sidings at more frequent intervals?

MC provided theoretical advantages.  These advantages are highly dependent upon the turnout size (bigger is better), signal spacing, signal aspects that can be used, and curvature and grades.

I like the advantages created by the mid-siding universal crossover which enables runarounds to occur outside of a terminal, and runarounds to occur without nearly as much delay to the following train because it gives the first train a long deceleration track.  It allows three-train meets to occur without heavy train delays and without committing the dispatcher to the final outcome nearly as far in advance as two separate plain sidings does.  It provides a place to kill a train that a terminal can't take and not lose the siding altogether.  It allows the track inspector or other maintenance of way to clear up for a train, without losing a siding, and without losing the time it takes to set off on a grade crossing.  That makes the track guys more productive and results in better track maintenance, and higher train speeds because the slow orders come off faster.

That's all theory.  Practice is different.  I couldn't possibly answer whether the actual installation is better than 2 or 3 separate 10,000' sidings because I haven't seen the line, studied the line, and learned why the railroad thought this was better.  It's quite likely that this siding design is quite specific to this particular line and its traffic. On another line with different geography and traffic it might be better to go with two 10,000 foot sidings spaced six or eight miles apart. 

If you're wonder if it's possible to study this scientifically instead of going by experience, it is.  The formulas were developed over 100 years ago.  The problem is, there are many variables, and at any rate, the output is 100% dependent upon what you think you're going to run for trains -- length, tonnage, type, priority, horsepower-per-ton, frequency, spacing, direction, bunching, seasonality, ad nauseum.  It's easy to build a model for a given set of train parameters but all of these variables can and probably will change on you in the real world and the perfect siding placement can a year later now be in completely the wrong place.

Worse, the planet isn't as smooth as a billard ball.  After you've decided what length of siding you need, you try to place it on the line and run into all sorts of grief trying to fit it around numerous natural obstacles such as rivers, curves, towns, grade crossings, and industrial spurs that get in your way.  The process is iterative and usually takes months to do, with input from a few dozen people that all have some particular expertise that bears on the decision.  The final location and length of a siding are often just the best that can be done for a reasonable cost. 

I do this for a living, among other things, so I can attest that the actual job of deciding siding location and length seems fun at first but by the time you're done and the track is laid, you wish you'd taken up a more relaxing job such as alligator taunting. 

RWM

[Modelcar]

...It seems such a waste that years ago many RR's had to pull up  double track {due to mostly economic reasons}, and now that extra capacity sure could be put to use. 

We understand why they were forced to do so, supply and demand of business and capacity, etc....Now it seems so unfortunate that it happened.

[anb740]

Not really.  With the longer siding (with 50mph turnouts), trains are able to meet opposing trains and keep up their speed without stopping.  And when you're trying to cram 40+ trains a day onto 200+ miles of trackage, timing is a big issue.  Case in point....here are two trains passing each other at 45mph on the new "Super Siding" at Lilly, GA. 

http://anb740.rrpicturearchives.net/showPicture.aspx?id=507198

Had this meet occurred on the original 6600' siding here, the northbound would have had to stop and eat a minimum 45 minute delay.  And three years ago before this expansion took place, crews were guranteed 12 hour workdays with up to two recrews just to get one train from Waycross to Manchester!

[mudchicken]

Try keeping the train in the hole in motion, albeit slowly, while the other train goes by. (running meet, less inertia to overcome)...everything stays fluid, and if the engineering planners are with it, a faster path to doubletracking.

[Murphy Siding]

     The article Eastern Corridors, in the Feb. 2008 issue of Trains Magazine notes a section of CSX track built with 'super sidings'.  They are 20,000 long, double track, with multiple crossovers.  Why would these be any better than have 2 or 3(?) shorter sidings at more frequent intervals?