Number out last Friday. CSX has plateaued. Train speed speed down a tick. Dwell up a tick. Cars on line up a tick.
All still below last May.
On the other hand, NS has been slipping and now has dwell and speed below CSX, and well below their 2013 levels when things were running smoothly. Over the years, NS has tradtionally been a tick better than CSX on speed and dwell.
-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/)
blue streak 1Oltmann: with both CSX and NS closing Chattanooga humps does that cause cross contamination for interchanges ? 1+1= -1 ?
Probably not. NS stopped doing car classification work there, but is still flat switching the local and interchange traffic as well as doing a lot of block swapping there. Chattanooga is a good, logical place for block swapping.
Hunter has changed trains on the A&WP sub. The trains have much more operating power than a few weeks ago.
Oltmann: with both CSX and NS closing Chattanooga humps does that cause cross contamination for interchanges ? 1+1= -1 ?
BaltACD oltmannd NS's numbers aren't anything to write home about. They still can't get the south end of the RR up to speed after closing the hump at Chattanooga. Running with one wheel on the shoulder.... Bold plans are great. As long as they encompass more than one 'bold action' and support the bold action with thousands of supporting elements. CSX and NS at Chattanooga appear to have made the bold plans of closing humps, without any of the supporting elements which makes the bold plan a humungos mess.
oltmannd NS's numbers aren't anything to write home about. They still can't get the south end of the RR up to speed after closing the hump at Chattanooga. Running with one wheel on the shoulder....
Bold plans are great. As long as they encompass more than one 'bold action' and support the bold action with thousands of supporting elements. CSX and NS at Chattanooga appear to have made the bold plans of closing humps, without any of the supporting elements which makes the bold plan a humungos mess.
Yep. Many issues, apparently.
oltmanndNS's numbers aren't anything to write home about. They still can't get the south end of the RR up to speed after closing the hump at Chattanooga. Running with one wheel on the shoulder....
Never too old to have a happy childhood!
This weeks STB posted numbers are out.
CSX basically treaded water Train speed and dwell off just a tick. Waycross really bad - holding cars for FL most likely. Cars delayed flat. Cars on line down a bit.
All things considered, not awful, but they are still in the ditch.
NS's numbers aren't anything to write home about. They still can't get the south end of the RR up to speed after closing the hump at Chattanooga. Running with one wheel on the shoulder....
BaltACD blue streak 1 oltmannd The good places to do block swaps often don't have yard air. They're more often just a "wide spot in the road". It's not do or die, just one more thing that needs managed. That could be of some expense. Electric, compressor, piping & probably at both ends of a siding ? $10k at least for each installation ? x 10 at least and probably closer to x 20.
blue streak 1 oltmannd The good places to do block swaps often don't have yard air. They're more often just a "wide spot in the road". It's not do or die, just one more thing that needs managed. That could be of some expense. Electric, compressor, piping & probably at both ends of a siding ? $10k at least for each installation ?
oltmannd The good places to do block swaps often don't have yard air. They're more often just a "wide spot in the road". It's not do or die, just one more thing that needs managed.
The good places to do block swaps often don't have yard air. They're more often just a "wide spot in the road". It's not do or die, just one more thing that needs managed.
That could be of some expense. Electric, compressor, piping & probably at both ends of a siding ? $10k at least for each installation ?
x 10 at least and probably closer to x 20.
Or use rental mobile compressor units and flexible hoses. That's what Uncle Pete did (and does) at certain points. Some places have since received permanent air plants, mostly at yards that originate trains. That way the car men can do the air tests without the power on.
A couple of places, one which has since received a complete air plant, are located outside yards where incoming trains would cut off the inbound power and take it to the house. The train is put on the yard air to maintain the air slip. (The air slip is placed in a mail box for the outbound crew.) The hostlers or outbound crew then put on the outbound power later.
Jeff
blue streak 1That could be of some expense. Electric, compressor, piping & probably at both ends of a siding ? $10k at least for each installation ?
Not to mention the ongoing cost of power to run it...
Larry Resident Microferroequinologist (at least at my house) Everyone goes home; Safety begins with you My Opinion. Standard Disclaimers Apply. No Expiration Date Come ride the rails with me! There's one thing about humility - the moment you think you've got it, you've lost it...
BaltACD1 - 2 & 3 are basic Old Time railroading
Well, sort of. One of the big changes that came from the scheduled operations rather than tonnage operations was the use of network models to optimize blocking and minimize handling. Before that, the classifications were decided by guys who "just knew". When railroads were smaller, this was possible. Not so much now with roads like NS and CSX.
Scheduing units trains is much harder than having schedules for merchandise trains because the RR and the shipper spend a lot of time not trusting each other. You don't win this game by just saying "trust me!" I have not heard any part of EHH's scheme that addresses this.
oltmanndThere is a snippet of a talk by a CSX commercial guy to a room of shippers that CSX posted on Facebook this morning. He explains PSR to the group. 1. Reduce intermediate handlings 2. Pre block at serving yards/long haul blocking plan 3. Make strategic en route set outs and pick ups to eliminate circuity 4. Don't force shippers into unit train service if cycle time is lousy 5. Integrate everything into the plan. MOW work, unit trains, etc. Items 1, 2 and 3 are basic tenets of scheduled railroading everywhere and have been for a decade. Item 4 is interesting but I doubt there is much "there" there. There have been times where marketing and operations have been at odds about service, i.e. discounts for large volume blocks drove lumpy traffic patterns. Item 5 is easier said than done. NS adjusts the plan for MOW gangs. Scheduling unit trains is hard work. Current efforts are less than rigourous efforts to "slot" them into the flow. There are probably some decent savings to be had if you can time the "launch" of unit trains around crew availability and track space. But, because 1-3 are already being done to a large degree, I'll stick to my original prediction that EHH won't get nearly the same sized operational "bump" on CSX that he did elsewhere. Most of the low hanging fruit was picked a decade ago.
1. Reduce intermediate handlings
2. Pre block at serving yards/long haul blocking plan
3. Make strategic en route set outs and pick ups to eliminate circuity
4. Don't force shippers into unit train service if cycle time is lousy
5. Integrate everything into the plan. MOW work, unit trains, etc.
Items 1, 2 and 3 are basic tenets of scheduled railroading everywhere and have been for a decade.
Item 4 is interesting but I doubt there is much "there" there. There have been times where marketing and operations have been at odds about service, i.e. discounts for large volume blocks drove lumpy traffic patterns.
Item 5 is easier said than done. NS adjusts the plan for MOW gangs. Scheduling unit trains is hard work. Current efforts are less than rigourous efforts to "slot" them into the flow. There are probably some decent savings to be had if you can time the "launch" of unit trains around crew availability and track space.
But, because 1-3 are already being done to a large degree, I'll stick to my original prediction that EHH won't get nearly the same sized operational "bump" on CSX that he did elsewhere. Most of the low hanging fruit was picked a decade ago.
1 - 2 & 3 are basic Old Time railroading
RE: 4 - Unit Train cycle times - who is the culprit for lousy cycle times? Shipper, Consignee or the carrier? A nonsensical statement from my vantage point
RE: 5 - When I was working the Operating Plan was changed on a weekly basis to account for and work around MofW curfews that were taking place at various places around the system. If necessary trains would be rerouted around work area, in some cases even rerouting trains over 'foreign' carriers.
EHH is not bringing anything new to CSX except terror.
Saturnalia oltmannd Pre-blocking is a wonderful thing and you only need a couple of hours planned dwell for a block swap, but there is only so much of it you can do and you have to run the trains on time to avoid having your air slip time out if you miss your connection. To what extent might pre-blocking save on car department forces? If you're in a hump you're going to Class-One every outbound airtest, whereas with blocking, so long as you have ground air, you in *theory* don't have to do a Class One ever, so long as the block swaps happen where your air is. Of course, new blocks would have to be Class-Oned, but if you hump your train, you start your air from scratch. It seems like most major yards these days have a significant amount of ground air, so I can't imagine that it'd be super expensive to extend the network in those cases where it doesn't reach to where you need it. But I have a hard time seeing how the departure yards in humps don't already have the air, in places where carmen make the airtest. If I have my airtest procedures wrong, by all means straighten me out! As to some of the former points, especially in terms of capacity for growth, I'd agree with the statement that eliminating a hump could make future growth more difficult. However, that'd only be the case in some places where traffic has the capacity to grow, and in places where there are enough different origin-destination pairs to make blocking inefficient, and where there are at least 1000 cars, if not 1500 cars per day, likely to pass through. I think there is little argument that CSX had too many humps - but in terms of getting it down to as few as two? Yeah I think that might be pushing it. 5-6 seems like a reasonable number to me. Willard, Queensgate, Waycross, Avon, minimum. Then pick your poison on which of the smaller ones to keep. Perhaps Selkirk or Nashville or Birmingham, to cover one or more of the "corners" not served well by those core four.
oltmannd Pre-blocking is a wonderful thing and you only need a couple of hours planned dwell for a block swap, but there is only so much of it you can do and you have to run the trains on time to avoid having your air slip time out if you miss your connection.
Pre-blocking is a wonderful thing and you only need a couple of hours planned dwell for a block swap, but there is only so much of it you can do and you have to run the trains on time to avoid having your air slip time out if you miss your connection.
To what extent might pre-blocking save on car department forces? If you're in a hump you're going to Class-One every outbound airtest, whereas with blocking, so long as you have ground air, you in *theory* don't have to do a Class One ever, so long as the block swaps happen where your air is. Of course, new blocks would have to be Class-Oned, but if you hump your train, you start your air from scratch.
It seems like most major yards these days have a significant amount of ground air, so I can't imagine that it'd be super expensive to extend the network in those cases where it doesn't reach to where you need it. But I have a hard time seeing how the departure yards in humps don't already have the air, in places where carmen make the airtest.
If I have my airtest procedures wrong, by all means straighten me out!
As to some of the former points, especially in terms of capacity for growth, I'd agree with the statement that eliminating a hump could make future growth more difficult. However, that'd only be the case in some places where traffic has the capacity to grow, and in places where there are enough different origin-destination pairs to make blocking inefficient, and where there are at least 1000 cars, if not 1500 cars per day, likely to pass through.
I think there is little argument that CSX had too many humps - but in terms of getting it down to as few as two? Yeah I think that might be pushing it. 5-6 seems like a reasonable number to me. Willard, Queensgate, Waycross, Avon, minimum. Then pick your poison on which of the smaller ones to keep. Perhaps Selkirk or Nashville or Birmingham, to cover one or more of the "corners" not served well by those core four.
There is a snippet of a talk by a CSX commercial guy to a room of shippers that CSX posted on Facebook this morning. He explains PSR to the group.
Paul_D_North_Jr How does the car get from Allentown to Pavonia ? Isn't that a lot of backhaul - Allentown to Reading to Philadelphia to South Jersey ? - PDN.
How does the car get from Allentown to Pavonia ? Isn't that a lot of backhaul - Allentown to Reading to Philadelphia to South Jersey ?
- PDN.
It's a bit of circuity. Allentown is in the wrong place in the network for current traffic. It would be better if it were in Reading. Allentown has been on the bubble for decades. I suspect if NS were to close Allentown, you'd see a lot of block swapping in Reading - provided the yard there could be beefed up sufficiently.
The circuity of Phila/South Jersey traffic from the west and south is the biggest problem of this kind on NS. (or was a couple of years ago, anyway)
An car trip example. A load of paper going from SE VA to South Jersey.
Local picks up car and takes it to Crewe VA, which is serving yard for the industry.
Car is classified "Allentown" at Crewe and picked up the next morning at 9AM by train 159.
159 drops Allentown block at Lynchburg VA.
6-1/2 hours later, 36Q picks up Allentowns and takes them to Allentown. This connection should be pretty good. It's not likely 159 will be too late into Lynchburg to miss 36Q.
Car is classified Pavonia at Allentown.
24-1/2 hours after arriving, it departs on 38G for Pavonia. This will be a very reliable connection. The inbound train can be 14 hours late and it will still likely make the right 38G.
Pavonia is the serving yard for load. It switches the car and builds the local train that departs 8-1/2 hours after arrival.
Total serving yard to serving yard trip time is 2 days 9-1/2 hours.
Car was had one intermediate classification at Allentown, and one block swap at Lynchburg.
Total dwell at Lynchburg and Allentown, 33 hours. Total time on a road train, 25-1/2 hours (if I did the math right...)
Serving yard to serving yard mileage, about 600.
This would be a slightly above average carload trip on NS.
Hump yards are a very efficient, cheap and fast way to classify cars. But, they are cheap only if you keep them busy.
It is expected that you can get a car from arrival to departing on the next train in 10 hours in a hump with great regularity. That yields an avg dwell of around 22 hours if all inbound and outbound trains run 7 days a week and distribution of cars "from" and "to" trains isn't lumpy (and generally, it is fairly uniform)
Pre-blocking is a wonderful thing and you only need a couple of hours planned dwell for a block swap, but there is only so much of it you can do and you have to run the trains on time to avoi- having your air slip time out if you miss your connection. One of the on-going activities in the service design groups is using the models to find block by-pass and pre-blocking. opportunities as the traffic slowly changes
The goal of your of you operating plan is to minimize the number-going of times you have to classify a car en route. The train service and train speed mean much less than the number of times you have switch a car. Roads like NS and CSX likely have it down to 1.5 to 2.0 times per trip (not counting serving yard or interchanges) with an average trip around 500 miles. That's pretty efficient.
CSX route structure and loose car network are not CP's or CN's or IC's. Square pegs and round holes
"If train A arrives at 8 AM with a mix of cars in it's 150 cars that need to depart on Train X at 12 Noon - that unless severe priority is given - the cars will make Train X tomorrow in the normal course of business."
However, if train A arrives at 8 AM with a block of 40 cars, it can set those out on a track ready for Train X to lift prior to its 12 Noon departure. Naturally the cars on Train A still have to be preblocked somewhere back on its route. That may mean a yard crew has to work a full shift rather than getting an early quit after 5 hours, and of course the arriving train crew has to do a little more work before booking off. But there is minimal additional labor cost and block swapping does seem to work fairly well now here in Calgary on the CPR. Because it works well in one place does not mean it will work as well for a more complex route network.
A hump requires a lot of maintenance to keep the retarders, power switches and other electronics working reliably, plus all the folks actually operating the humping process on the ground and in the control tower. Yet more yard crews are needed to move cuts from the class tracks over to the departure tracks to assemble a future train.
Modern unit trains and intermodals bypass hump yards and there is a lot less loose car railroading these days. Why the railroads have tried to discourage the latter is separate topic, but other threads have discussed how the railroads seem to ignore obvious business opportunities.
With or without Hump Yards - for the most part there is daily service between Origin-Destination pairs. That is a 'scheduled' 24 hour delay to the car that, for whatever the reason, misses today's train. The reality of train operations at virtually any yard (hump or otherwise) If train A arrives at 8 AM with a mix of cars in it's 150 cars that need to depart on Train X at 12 Noon - that unless severe priority is given - the cars will make Train X tomorrow in the normal course of business. To make a 12 Noon departure, Train X needs to have it's switching completed no later than 9-10 AM - thus giving the Car Department time to lace the air hoses and make a departure Class 1 air test - identifying shops and making any 'running' repairs that they are equipped to perform - while this inspection is taking place the track(s) Train X occupies will be Blue Flagged.
Flat switching does not increase switching efficiency. It eliminates the cost of maintaining the hump retarders and power operated switches and computer support of the operation - at the resulting increase in dwell and car hire as well as facilities being occupied for a longer period of time.
Paul_D_North_Jr oltmannd Numbers for the past week (STB724) are out. Train speed is up over 1 mph lead by intermodal speed. Cars on line are down about 4k. Loads and empties over 48 hrs are down a nice chunk. Dwell is the sticky one. Didn't really budge. Maybe because of traffic held over Labor Day? The numbers showed the improvment that can happen when you use a slack period to do a 'reset'. Do they have enough resources to hold these numbers and make more improvement? Might take a few weeks to find out with Irma disruption... On Labor Day (!) I saw or heard about a half-dozen CSX trains pass through Ohiopyle, PA in the 9 AM - 7 PM timeframe, though I was out of earshot from the tracks from about 2 PM to 6 PM. So despite the usual almost-shutdowns on major holidays, CSX was pretty active. FWIW, the 2 trains I actually saw were a double-stack and a multi-level, and 1 of the moves I heard could have been just a light-engine helper set. - PDN.
oltmannd Numbers for the past week (STB724) are out. Train speed is up over 1 mph lead by intermodal speed. Cars on line are down about 4k. Loads and empties over 48 hrs are down a nice chunk. Dwell is the sticky one. Didn't really budge. Maybe because of traffic held over Labor Day? The numbers showed the improvment that can happen when you use a slack period to do a 'reset'. Do they have enough resources to hold these numbers and make more improvement? Might take a few weeks to find out with Irma disruption...
The numbers showed the improvment that can happen when you use a slack period to do a 'reset'. Do they have enough resources to hold these numbers and make more improvement? Might take a few weeks to find out with Irma disruption...
On Labor Day (!) I saw or heard about a half-dozen CSX trains pass through Ohiopyle, PA in the 9 AM - 7 PM timeframe, though I was out of earshot from the tracks from about 2 PM to 6 PM. So despite the usual almost-shutdowns on major holidays, CSX was pretty active. FWIW, the 2 trains I actually saw were a double-stack and a multi-level, and 1 of the moves I heard could have been just a light-engine helper set.
When you are in the ditch, you use the holiday to try to get out and run everything you can. When things are good, but traffic is slipping, you try to lay down everything you can. In normal times, it's usually a blend.
daveklepper Sat: I do not see how converting hump yards into flat switching yards benefits loose-car railroading. Hump yards can take four inbound trains from different points and convert them into four outbound trains to different points a lot faster than a flat-switching yard.
Sat: I do not see how converting hump yards into flat switching yards benefits loose-car railroading. Hump yards can take four inbound trains from different points and convert them into four outbound trains to different points a lot faster than a flat-switching yard.
DaveK,
Yes, hump yards can do that with 30 hour or so average dwell, BUT hump yards are expensive so the object is to hump at as few places, or as few times as possible in any given trip. You need to think in blocks of cars not individual cars. Any yard builds or breaks down blocks of cars. If I can preblock to avoid a hump yard down stream I have probably cut costs. Now perhaps my thru train simply needs to set out a block of local cars for the former hump and pick up another block or two or three or four blocks that were flat switched. Some switching can be done here or there. Which is best, cheapest, quickest? This is almost impossible for an outsider to answer and is why the railroads have operations models that can test different plans, determine the best, and locate bottlenecks which may start an investigation of alternatives and costs to eliminate the bottleneck or relocate the work.
What you do not want to do is change a bunch of stuff all at once, expecially if you have not run your model. Others have talked about just making changes and seeing how it will work out. That is OK in moderation, after all the model may not be correct. The key is moderation and doing a series of changes incrementally so they can settle and you can see what the result was and prove or disprove the model.
It appears that EHH has simply issued orders based on his private theory of what will work. He could be right. Done incrementally if something doesn't work, it can be identified and reversed or modified. If you change a whole bunch of things all at once it is a formula for chaos. If EHH's vision is correct, the chaos will clear up. If vision is not correct, it will be tough to sort out the good changes from the bad. So far no one on the outisde, and perhaps on the inside too, knows if EHH vision is correct.
Mac
Sat: I do not see how converting hump yards into flat switching yards benefits loose-car railroading. Hump yards can take four inbound trains from different points and convert them into foiur outbound trains to different points a lot faster than a flat-switching yard.
The coal trains in general did not require the hump yards as much as general manafest trains. And both NS and CSX do have important lines capable high speed if the track, signals, and dispatching permit, New Jersey - Florida for example.
I think in general your reply more substiantiated what I posted rather than contradicted it. If HH does in fact go after new business and gets it, than CSX won't have to dump him to recover. I hope this is truly the case.
cx500Major holidays are irrelevant in the Precision World of EHH, just a nuisance and another way to harass the running trades.
Over the past several years the only Holidays CSX 'planned' for were Thanksgiving, Christmas and New Years. Work during other 'holidays' were dictated by how the customers observed the holiday. If the customers for a particular job didn't work for the holiday, that job would not work for the holiday. Operationally, very few industries observe more than the traditional three - their backoffice and adminstrative positions may observe the holiday but operations continue normally.
Major holidays are irrelevant in the Precision World of EHH, just a nuisance and another way to harass the running trades.
oltmanndNumbers for the past week (STB724) are out. Train speed is up over 1 mph lead by intermodal speed. Cars on line are down about 4k. Loads and empties over 48 hrs are down a nice chunk. Dwell is the sticky one. Didn't really budge. Maybe because of traffic held over Labor Day? The numbers showed the improvment that can happen when you use a slack period to do a 'reset'. Do they have enough resources to hold these numbers and make more improvement? Might take a few weeks to find out with Irma disruption...
Our community is FREE to join. To participate you must either login or register for an account.