THE ONE-SPEED SCHEDULED RAILROAD

[kevarc]

Mailline fueling reduces terminal time for hot trains. The average time in a yard for engine changes or removing the engines and service them and return them to the train has got to be a lot longer than mailline fueling. IT really depends on how the refueling station is set up.

[Junctionfan]

Can't they build special sidings for that than instead of delaying a whole bunch more trains? Me personally, I would build staging yards for fueling and maintainance as well as eating and rest facilities for the crews like a truck stop for trains, (service station, diner, 7-11) like atmosphere with tracks instead of tarmac.

[kevarc]

Would be like driving along the turnpike! That would actually make a bit of sense so they probably will not do that.

IIRC, the fueling station are on seperate tracks, they still have bypass tracks around the fueling station.

[Anonymous]

QUOTE: Originally posted by Junctionfan

Wouldn't it be cheaper if the railroads payed the crews by an annual salary than? Would the workers / unions go for that?

One way compromise is to pay the crew for the trip regardless of artbitraries. A straight trip pay assuming it is greater than would have gotten normally would give the crew an incentive to get the train over the road as quickly (and safely) as possible and not dog it in order to collect overtime and "artbitraries" . Artbitraries is the addtional pay that the RR has to pay the crew for conditions and circumstances that prevented the crew from bringing in the train in ontime like outlawing on the Hours of Service law. This varies from local to local.

I talked to a RR executive who work for the Southern Pacific subsidary SPCSL Lines when they bought the old Chicago and Alton. He said the crews really moved those trains quickly with one crew between Chicago and St. Louis.

[Junctionfan]

On my railroad I would give out salaries with bonus for crews who get the trains ahead of schedule even if it means a few seconds. That way they still have the assentive to try to avoid assentives but won't be penalized for any arbitraries that may have occured on their shift.

Of course I'm one of those sorts who believes that you should treat your workers like your Rolls-Royce (if you're a big wig executive who owns one)

[TH&B]

Another way to increase avaerage speeds at terminals is to have block cicuts in receiving tracks and pad tracks and other improvements so that trains won't have to enter at "restriced" speed where under circumstances like poor visablity in bad weather or curves with a short line of sight can force engineers to move well below the 15mph.

If it takes abiut 15 mins to pass a 2 mile train at 8mph wich can sometimes be a reasonable speed to go at restricted speed you are realy reducing capacity and if you have to cross over to get into the yard then you are blocking both mains for all this time.

[daveklepper]

On another thread, the mistaken idea that speeding up auto traffic improves capacity was stated. I had to note that the optimum speed for private auto traffic past a specific point is 22 mph, because of the terrific increase in stopping distances as auto speed is increased. That got me thinking about the ideal speed for a one-speed railroad in terms of capacity. 70 mph is too fast. When stopping distance is less than the train length, then addtional speed ads capacity. What is optimum or close to it, is a stopping distance equal to train length. When the stopping distance is much longer than train length, higher speed means less capacity. I would there surmize that 45-50 mph is a good speed for capacity, where a one mile long train takes about a mile to stop on level track with an emergency but controlled application. Under these situations, with normal block signal practice spaced say a bit more than every 1/3 mile, for four indication signals, we should be able to pass a train about every 10 minutes safely, and assuming no window for the junk train or maintenance, say the one heaviest traffic day of the week, that means 144 trains handled in the one direction. This can be doubled to 288 trains with really modern signalling and ATC, where you have a sliding block, a signal system that sort is a radar guided by the rails that measures the distance up to say ten miles to the train ahead if any and is coordinated with 4-postion cab signals and automatic train control. We should then be able to run trains every five minutes. But I agree that terminals, yards, are the real big problem, and that is why the original title included "scheduled." These are all idea that might influence the future profitability of the big six if applied with good reason in the areas that are applicable.

[Overmod]

Don -- with GREAT respect, the purpose of modern motive power is to be moving in revenue service as many hours of the day as possible. The more time power spends cut off a revenue train, the greater the opportunity waste. This is even more true for unit trains than for 'mixed freight' power that is idle between arrival at a yard and departure with the next made-up consist.

There may be a potential gain with 'on-line' fueling at division points of a run-through train, particularly one with a long route. It takes fuel to carry fuel, and in other industries (admittedly for other reasons) you often find that operating procedure dictates running with less-than-full tanks at times. It would not make sense to accomplish refueling of a made-up consist in transit by cutting off power, moving it to a service rack, etc. each time, particularly if only a 'partial' load of fuel has to be transferred and the on-line fuel stop can be coordinated with other required stoppages of the train (e.g. Government-mandated distance between inspections of the train, or crew changes)

I think we're all agreed that it is NOT optimal to have a required fuel stop directly on a critical piece of main. I would wonder whether 'fueling sidings' might make sense if coordinated properly, or whether containers or cars of fuel could be spotted at strategic locations on the road to permit 'topping up' at times the train is held for valid operational reasons.

[oltmannd]

Agree completely with this exception.

On Conrail, you could count on being able to go 1300 mile minimun w/o needing refueling on intermodal trains, so, theoretically, you would never need main line fueling. Other factors intervened, however. At both NJ and Chicago, you had multiple terminals, so you would have needed multiple fueling stations. The old installations had leaky steel tanks that had to be removed and were costly to replace. Also, in NJ some of the areas were environmentally protected, so fueling couldn't be installed. Direct from truck is possible, but slow and expensive. Finally, fuel cost and tax figure on where to fuel. e.g. Chicago costs more than Elkhart. Harrisburg is cheap because it's adjacent to Colonial pipeline terminal.

Still, I'll stand by my "stupid" comment. It's always a bad idea to build in a bottleneck. Do the things that take time when you have time, not when you could be on the move.

[oltmannd]

QUOTE: Originally posted by Overmod

Don -- with GREAT respect, the purpose of modern motive power is to be moving in revenue service as many hours of the day as possible. The more time power spends cut off a revenue train, the greater the opportunity waste. This is even more true for unit trains than for 'mixed freight' power that is idle between arrival at a yard and departure with the next made-up consist.

There may be a potential gain with 'on-line' fueling at division points of a run-through train, particularly one with a long route. It takes fuel to carry fuel, and in other industries (admittedly for other reasons) you often find that operating procedure dictates running with less-than-full tanks at times. It would not make sense to accomplish refueling of a made-up consist in transit by cutting off power, moving it to a service rack, etc. each time, particularly if only a 'partial' load of fuel has to be transferred and the on-line fuel stop can be coordinated with other required stoppages of the train (e.g. Government-mandated distance between inspections of the train, or crew changes)

I think we're all agreed that it is NOT optimal to have a required fuel stop directly on a critical piece of main. I would wonder whether 'fueling sidings' might make sense if coordinated properly, or whether containers or cars of fuel could be spotted at strategic locations on the road to permit 'topping up' at times the train is held for valid operational reasons.

In order to get off the main at 40 mph, you'd need a 4 mile approach siding, and I suppose, even more to get back on the main at 40 mph. That's a lot of expensive track and signalling.

I agree it's better to fuel power on a train then to cut off and go to the engine house, but that's rarely the only alternative. You would be shocked at how many times trains fuel on the main that could go all the way to destination w/o adding any fuel. Truth is, if the train passes the rack, they top it off. Since there is no "fuel plan" for the locomotive, they top it off every time, just in case. Many, many times they're adding only 1000 gal or so to a 4000 or 5000 gal tank. RRs could get a lot smarter about where and when to fuel.

[kevarc]

"On my railroad I would give out salaries with bonus for crews who get the trains ahead of schedule even if it means a few seconds. That way they still have the assentive to try to avoid assentives but won't be penalized for any arbitraries that may have occured on their shift."

Good idea, but the odds are that it will not happen. Why? Union Contracts. When I was in the UMW, the conpany wanted to give bonus to crews the beat their goals. The knew they couldn't give money because of theunion contract. So since they were owned by an oil company, they wanted to give motor oil, etc. Guess what? The union folks stepped in and said no way - either everyone gets it or no one gets it. So no one got it.

This is going to drive me crazy - maybe Mark will remember who it was. A while ago, I want to say in the 80's, UP (I think) got upset with fuel taxes in one state that they set up a fuel station before the state line so they would not have to fill the engines in that state.

[Junctionfan]

I imagine the more grades the train encounters, the quicker it is for the trains to require a fuel stop as well as possible sand and oil refill. On my pretend railroad, there isn't too many grades to worry about until you get between Boston and Rotterdam Jct NY on the Guilford B&M line so I can run a train between Chicago and Buffalo no problem without worring about fuel but past that, I think I would need to do some system checks if you know what I mean.

That is probably why UP and BNSF will run 70mph intermodals on the flat planes but slow down on the hills not withstanding that it could be because the curves might be sharp and there could be debris on the tracks from rocks and fallen trees around the mountain side.

[Anonymous]

Don't forget that track speed is not a made up number. The FRA has several classes of track, with speed limits attatched to each class. The higher the class of track, the higher standards of maintenance, the higher the speed limit. This of course, means a higher cost.

[Paul Milenkovic]

M. W. Hemphill gives numbers that convert to the "gas mileage" of a single 100 ton hopper car as part of a train as 19.5 MPG at 30 MPH and 9.7 MPG at 60 MPH. Oltmannd gives the gas mileage of a loaded 100 ton hopper as 32 MPG at 30 MPH, 20.4 MPG at 60 MPH -- for an unloaded hopper the figures are 69.2 MPG and 40.9 MPG.

M.W. Hemphill bases fuel consumption on so many SD70MACs in a consist running at Run 8 to maintain track speed -- the numbers of fuel consumption for different locomotive types can be found at http://www.vcn.com/~alkrug/rrfacts/fueluse.htm. The numbers seem pessimistic. Is Run 8 required to maintain speed on level track? I had always thought you needed reserve to accelerate to speed and to climb the small gradients that occur even apart from major climbs. I also have read in trucking trade journals that a modern 18-wheeler can get about 6 MPG at 65 MPH, and 9.7 MPG for the 100 ton hopper at 60 MPH, and given the rolling resistance advantage of rail and the air resistance advantage of freight cars of identical profile close-coupled in a train, that number seems low.

Oltmannd gives numbers based on over-the-road experience, and they are in the ballpark of Davis formula values for rolling resistance and a drag coefficient (C_d) of .2 for a hopper car in a close-coupled train of identical cars of the same profile (TOFC has horrid drag numbers because of the gaps between trailers). My numbers suggest that at 60 MPH, about half the drag is aero, half rolling, which is pretty much confirmed by the loaded/unloaded difference.

M.W. Hemphill gives a price on a hopper car of about $50,000 -- do I have this correct? That number seems low. I thought rail cars cost $50,000 back in the days that locos cost half a million. I am thinking that with locos in the 2-3 million and up range, a hopper car must cost a cool $200,000. I am also thinking a 100 ton aluminum hopper weighs about 40,000 lb, about half in steel running gear and half in aluminum. The materials alone must run $50,000 before you take into account fabrication.

I am thinking that if you boost the car price and use Oltmannd's fuel consumption numbers, 60 MPH is about where you should run a coal train -- maybe 70 is too fast but 30 MPH is way too slow. And the turn-around times don't matter because I am assigning the lease costs per hour that the train is running -- if the lease cost per hour is more than the fuel cost per hour out on the road, you are running that train too slow.

Back in the 60's when Integral Train Systems was published, Kneilling thought that coal trains should run at passenger train speeds -- 79 MPH, but oil was perhaps cheaper in relative terms then. He had numbers in the book to back that up.

Were the railroads completely stupid and John G. Kneiling an unrecognized financial wizard? In a way yes. Kneiling's point was that railroads were basing their operating decisions based on accounting formulas which gave unrealistic lifetimes to equipment (30 plus years) and unrealistically low levels of return on capital. The whole premise of Integral Train Systems was that railroads had too much capital sunk into rolling stock at low rates of return on account of unrealistic accounting formulas together with the business-as-usual approach to getting that rolling stock over the road.

[dehusman]

All this economics is good, but have you figured in the cost to rerail the cars when they derail at higher speeds?

The speeds on the track and cars are put there for a reason. The geometry of the track or the dynamics of the cars won't allow higher speeds. Saying that ou want to run a 19000 ton coal train at 70 mph is great, but if its not safe from a train dynamics standpoint then it doesn't matter. Saying you want to run a coal train at 70 mph when the curves won't permit speeds over 50 won't do you much good. Saying you want to run a train at 70 when the grades won't permit a train operating at that speed (braking power).

Heading into a yard at 40 mph is way cool except if you have to stop in that track to change crews or yard the train you are going to be on approach at best (30 mph) and slowing immediately down to 10 or 15 mph in preparation to stop at the other end, so spending all that money on an inbound train hasn't really gained you anything. The only place a 40 mph track and switches gets you a bang is on departure where a train can accelerate out of the yard track at 40 mph.

Dave H.

[daveklepper]

Isn't the way to remove terminal delays and refuling delays and power shortages a SCHRDULED railroad. Scheduled by people who know the equipment, know the traffic, and know the right of way? So that four trains from three different directions don't arrive at a receiving yard or fueling station all at the same time?

Regarding a one speed railroad. I believe this is what the N&W did at least in the steam days and probably Roanoke - Norfolk today . Not 70mph by any means, although there is that photo of an A pulling coal at 70, but fast enough for the merchandise and slow enough for economical coal haulage. Was it also a scheduled railroad?

[dehusman]

The trick with the scheduled railroad is balancing a network trains on a network of routes, each route with a different transit time, with different dwells at different terminals and different committments on different traffic and still keeping "even" slotting on trains at every terminal. Throw in that customers don't load on an even basis (loadings tend to increase towards Thurs-Fri) and so train volumes vary by day of week (5 trains on Monday and 8 trains on Friday) and vary across the country as the distance from the terminals to the major origin locations.

Railroads do attempt to schedule/optimize both the inbound and outbound scheduling from their yards.

The really cool thing to watch is after you get all this done, some drunk will weld his car to the front of a train and delay a dozen trains for 2 or 3 hours on one route while a thunderstorm reduces speeds while the track is patroled in another. In a short time reality pretty well scrambles your intricate scheduling. 8-)

Dave H.

[TH&B]

Yes removing leval crossings and fencing off the right of way would probably also need to be done to reduce chances of delays. The weather you cannot do anything about obviously but you gotta reduce as many unpredictable variables as possible.

I still don't see the need to travel over 45mph because we are raising the average speed by removing bottle necks.

[oltmannd]

Two items:

Check the NS web site today for the analyst presentation. In there, you'll find NS paid about $0.87/gallon for #2 in the 3rd qtr. They have a fuel hedging program, which smooths out the bumps and renders closer to a long term avg price for fuel, so $1.40/gal may be a bit high.

2. I'm sure we've been over this before, but trying to get loaded coal trains over 60 mph means either longer signal blocks or empty/load braking. The braking ratio of loaded hoppers is too low to get stopped many places. On Conrail, 50 mph was the fastest you could go and still make safe stopping distance.

[Grinandbearit]

All this discussion and no one has mentioned what happens when a 90-100 mph passenger train ( which takes up space equivalent to 3 or 4 freight trains) is thrown into the mix.

[oltmannd]

QUOTE: Originally posted by Grinandbearit

All this discussion and no one has mentioned what happens when a 90-100 mph passenger train ( which takes up space equivalent to 3 or 4 freight trains) is thrown into the mix.

That would be akin to the passenger train having a huge plow and just plowing thru the rest of the traffic!

Actually, if you have a single track line and you can get the passenger train sponsor/operator to double track it for you as part of the condition of operating at higher speeds, it's a win-win. That's a big IF, though.

[dldance]

I don't know anything about scheduling railroads - but I do know a lot about scheduling global supply chains in the electronics industry and I have learned a few rules that seem to apply in the context of this thread:

1 - There is always a bottleneck! In a capital effective system that bottleneck should be the most expensive issue to resolve. I have seen poorly managed systems where cheaper nodes became the bottleneck because someone was managing local utilization instead of system throughput - thus they prevented the most expensive node from being effectively utilized.

2- Murphy was an optimist!

3-The enemy of higher output is variability. (See rule #2) If all of those trains had exactly the same operating and braking characteristics - scheduling would be significantly simplified. That is why FedEx does such a booming business - the promise of overnight delivery eases the burden of manufacturing variability. (Don't think I am suggesting that we FedEx coal).

I have learned a lot from this discussion that I can apply in my world.

dd

[TH&B]

You don't through a 90+mph passenger train in, that is all there is to that or else we'll through a 45mph 18,500 ton train with 1 hp/ton onto the HSR.

I heard about a freight train in England derail while coming to an emergency stop in a very super elavated curve for high speed trains, the freight cars were apparently too top heavy !

[rob_l]

Mark,

Many thanks for a very illuminating analysis of unit coal train costs. It helps us to understand, for example, the transition of motive power for BN’s coal trains from 5 SD40-2s/C30-7s a decade ago to 3 SD70MACs (running more slowly) nowadays.

If I may, I’d like to append to the analysis a couple of considerations that tend to pu***rain speeds up a tad from the “optimal” speed. These considerations are (1) a delineation of costs into investments vs. the avoidable operating costs per trip, and (2) consideration of revenue, not just cost.

As we all know, one typically cannot “hire” coal cars and locomotives for each trip, one has to invest in car and locomotive fleets and pay for them whether they are used or not. And this changes things a bit. For ease of illustration, suppose we are a small road with one unit coal train customer. Suppose at the optimal speed, we would require 8.5 sets of equipment to cover the customer’s move at the forecasted volume level.

Now we don’t want to run a half-size train (pushes up labor costs) and we don’t want to run a train set half a year (because we have to pay for it all year), so we won’t buy more than 8 sets of equipment. Then we have to consider the investment risk. If the economy drops off or if the price of natural gas moderates, our coal customers’ business will go down, in that case we might need only 7 sets of equipment. Or maybe 6. Worse yet, maybe there is a risk the coal mine will lose its contract completely. (Remember New Mexico coal that used to go to Wisconsin Electric?) How many equipment sets we are willing to buy depends on our appetite for risk (and on the contract we can negotiate with the shipper). Let’s say for argument’s sake that we buy seven.

GIVEN we buy only seven sets of equipment, suppose our customer’s business is good right now. In fact, good enough to keep 8.5 sets of equipment busy at the “optimal” speed. Now, the consideration of revenue comes into the picture. We make a lot of money on unit coal trains; the revenue per trip is dramatically larger than the avoidable costs (fuel, labor, maintenance) per trip. If we speed up from the “optimal” speed, some of the avoidable costs will increase significantly (especially fuel), although some avoidable costs may moderate (e.g., T&E labor). Even if the percentage increase in avoidable costs is higher than the percentage increase in revenue, as long as the amount of revenue increase is larger than the cost increase, it pays to speed the trains up. Moreover, we should keep speeding up the trains until either (1) the increment in cost per year matches the increment in revenue per year, (2) we have captured all of the traffic demand, or (3) we don’t have enough locomotives to go faster. In this unrealistically small case, it pays to speed up the trains about 20%, if we have the locomotives and revenues to do it.

If we don’t speed up, we are leaving money on the table.

Now let’s think about a more realistically-sized case. As the number of shippers sharing the car and locomotive fleets grows, the percentage impact of rounding down to an integer number of equipment sets moderates. And if there are some offsetting risks of volume fluctuations at different customers that we serve, the risk of underutilized equipment may be diversified down a bit (because of the portfolio of customers). But the investment risk never goes away completely, and we will always be averse to buying enough equipment to handle the forecasted business at the “optimal” speed (because there is too big a chance of having cars and locos with nothing to do). Given the investments we do make, if the business is out there to utilize the equipment sets more, it always pays to speed up somewhat from the “optimal” speed and capture the additional revenue, provided our marginal cost is less than the incremental revenue.

And this is why the actual speeds we observe the RRs making are somewhat higher than the “optimal” speed.

Best regards,

Rob L.

[daveklepper]

Certainly on a scheduled railroad, the schedules should be demand responsive with respect to trends by responding both in frequency and in speed. It does require savey to run a scheduled railroad. Schedules have to be familiar with the territory, the business, and the possibilities for "events". And of course the equipment and the people. But I think it is worth doing.

[MP173]

FedEx running unit trains? Perhaps in a limited manner, but I what about the assembly and distribution aspects? Hardly unit train in manner.

The line haul aspect once the shipments get into the air could be thought of as unit train, but that is only a fraction of the entire transit time. Fedex kinda reminds me of the old REA business, except they handle complete movement.

Fedex also had various levels of service from NDA to ground, as does UPS, although Fedex has taken it to the next level with the LTL service.

Mark, do you know, does FedEx use intermodal in the manner UPS does for ground service?

ed

[Overmod]

Experts probably know for sure, but I think I do too. Not yet.

FedEx *has* considered both intermodal and bimodal for a number of runs and services, but has not embraced any of the technology yet to my knowledge. With respect to UPS-style high-speed block movements, they're waiting until UPS pays for the experimentation and finds all the teething troubles and logical inconsistencies. Once it works definitively, and some of the capital costs have come down with experience, volume production, and amortization/depreciation, they may participate. Certainly there is interest, both on the express side and the ex-Roadway ground side.

I don't know if you include Custom Critical as LTL, but their operational model there leaves very little advantage to any current form of intermodal/bimodal (except perhaps something like BladeRunner, which existing railroad arrangements would likely never tolerate at acceptable expense...)

[Anonymous]

Fascinating topic :)

Imho american railroads are not really structured for high speed freight (like 60 mph across the board). The most important factor would be the shortage of power.

Diesel locos, while great luggers, are rather lacking in power department. Adding 2 sd70macs to a pair of these to achieve higher speed will add extra maintenance - that is extra shop crews, extra shops and higher investment in motive power (locomotive prices are roughly per unit, not per hp).

What is needed to increase power from 2 sd70mac is not 2 extra sd70mac, but a pair of sd140mac ;) to replace the current pair of sd70mac. Since 8000 hp diesels are not yet avalible (and I doubt if they will be in any forseeable future) this would mean what american railroads detest...

zap-zap - electrification...

With current technology 12000 hp 6 axle electric loco is standard equipment, or would be if anyone wanted such monster.

But that would mean a major capital investment, one that would pay off in lower energy costs, higher power of locos and possibility of regenerative braking. But none of the railroads has several billion to string the wire... So that is also impossible.

Another problem with higher speed is increased track wear - this would either mean increased maintenence or reduced track life. One of the reasons why TGV weigh just 37800 lb tons per axle is to keep track maintenance costs in check.

IMVHO the cheapest way to increase capacity of a railroad is not to increase speeds but to improve signalling and braking to reduce headways. For instance ABS like system for all cars would allow the train to stop at shorter distance (ahd would save a few $ because flat spots would happen much less often).

[Junctionfan]

My idea of improved courier service is roadrailers leased to UPS, Purolator, DHL, FedEx, and ICS Courier. Load them all up and run them with a P-42 at top-speeds.

You wouldn't believe how much money people are willing to spend for "first class" service for courier service so maybe the railroads should attempt to use the demand better to their advantage.

I know that ICS has their own 40 foot containers BTW.

[dehusman]

If you want fast courier service, don't use a P42, use a B747.

Dave H.