Passenger Trains > Allow higher speeds to make up time on Western Railroads

Granted, there are many problems due to lack of crews and operating/handling of freight trains. But it has been a long time that Amtrak has really looked hard at the speeds for its passenger trains on the freight railroads. In earlier days Amtrak (I believe Jim Larson) would take its Corridor clipper or other track geometry car, bring freight railroad engineering folks aboard, and then negotiate and work with them to raise speeds on curves and other pieces of track. One example was the Southern Pacific that stubbornly refused to raise its top speeds from 70 mph until Amtrak worked hard with them to increase it to 79 (at least in some places). Another was the new Amtrak San Joaquin service that started with a maximum of 70 until Amtrak worked with hard the Santa Fe to increase the speeds on the Valley line to 79. Another was the Pioneer train that had maximum speeds of 70 until Amtrak worked closely with the Union Pacific to increase the speeds to 79.  Another was the Santa Fe that had the same curve speed restrictions for freight or Amtrak passenger trains below 55 mph (so that they wouldn't have to add an extra column in the timetable). Amtrak worked with  them in the 1980's, I believe, to allow for higher curve speeds for passenger trains. 

Many of the legacy passenger operations have never had their speeds adjusted because Amtrak has not pushed for them.  Granted, sometimes money is involved to re-time signals and gates, but many times it is not needed due to newer variable speed crossing gates that are now used and longer 2 mile signal blocks.  Sometimes it may be as easy as to put up new speed boards and a timetable. Under PTC requirements, maybe even calculations for new braking requirements by the system would not be required as restrictions to brake might be removed. 

It is amazing the difference that can be had with a little analysis and work. On the ex-Southern Pacific now Union Pacific Coast line between Newark and San Jose California, the speeds were usually only 60 mph for passenger trains from the days of the S.P. Oakland Lark. S.P. had raised speeds in a few places to 70, but it was a slow line, overall. The Capitol Corridor/ACE train management finally increased speeds on the line to 79 with many of the curves now at 75 mph with no line relocations an no severe curve banking. This was all done after PTC had been installed on the line. 

With PTC there is now no requirement for Automatic Train Stop (ATS) or cab signals to run above 79 mph, but in nearly all cases, no changes have been made to increase speeds. 4 quadrant gates are not even required until higher class 6 track standards or above. Welded rail has replaced less maintained jointed rail in many places with no increases to speed. In another instance, looking at the employee timetables for the BNSF in New Mexico and Arizona, they have removed the requirement for ATS, but have not increased the speeds on the other track or the other direction and 90 mph higher speeds are still only allowed in many cases on one track in one direction. There are also many places on Western Railroads where there is straight track maintained to class 5 standards (allowing 90 for passenger and 70 for freight) where no changes have been made to actually allow for running at such higher speeds. 

Increases to speeds without tightening the schedule would allow for the passenger trains to more easily keep to their schedules and allow for recovery from unforeseen consequences. Making these changes would not eat up a larger track slot but would allow for a passenger train to get out of the way quicker when there was no other train ahead. As a good example, I recall being on the Southwest Chief in New Mexico. This busy railroad found our train stuck behind a freight train going slowly up a steep grade. We proceeded for miles at 15 mph until the crossover.  We did not delay the freight. At that point we passed the train and then proceeded at the 90 mph allowed here and were very quickly out of the way of all other freights at Dalies.  We were quite early into Albuquerque even with this slow running behind that freight. If we had been more limited in speed we could have been in the way of other trains that were ready to go on this busy railroad. 

My call is to Amtrak (and others) is to get busy and work with the freight railroads for reasonable speed improvements as one way to to allow for more recovery time from unforeseen circumstances. This has been done in the past and it can be done again now. 

Amtrak doesn't set the speed limits, the owner of the track does. And they are determined by which standard the track was built and maintained to.

PHall Wrote:
-------------------------------------------------------
> Amtrak doesn't set the speed limits, the owner of
> the track does. And they are determined by which
> standard the track was built and maintained to.

This is acknowledged in the posting, but there's an implication that the passenger-train speeds on certain track segments could be increased essentially at no additional maintenance expense because the limits the freight railroads are, in essence, arbitrary (and not supported by the tracks' FRA maintenance classification) and have gone unchanged for extended time without reconsideration.
 

Just look at how the 110 mph on the Michigan Line
has improved timekeeping and how the 90 mph on
the Chicago-St. Louis line has resulted in faster schedules......

Not asking for improved schedules yet, just faster speeds to allow for more recovery when there are problems. 

Superelevation isn't cheap.  Not just to install but maintain as freights are much heavier than in the old days.
john

Having ridden the Empire Builder recently, very little if any of that track seems good for more than 79 mph, and the rough spots are pretty jarring even at that speed. Much smoother through the slow sections through the mountains.

Commuter/corridor trains can gain ridership by shaving off a few minutes. I could personally care less whether my journey from Portland to St. Paul takes 40 hours or 33 hours. Much more important to focus on repairing and replacing equipment, hiring adequate staff, and improving reliability.

dendroica Wrote:
-------------------------------------------------------
> Having ridden the Empire Builder recently, very
> little if any of that track seems good for more
> than 79 mph, and the rough spots are pretty
> jarring even at that speed. Much smoother through
> the slow sections through the mountains.

Between about Glasgow and Minot in a Superliner was like riding a caboose.

TAW

Increasing passenger speeds won't improve much unless freight speeds also increase. Otherwise you'd only create more conflicts with freight traffic. I agree there are many locations where passengers speeds could be higher, but most would not save enough time to justify the cost.

As for the upgrades mentioned on the UP Coast Sub in Norcal, the state paid handsomely for these upgrades which actually only save a couple minutes of running time, hardly worth the investment. Meanwhile very slow curves and turnouts still remain on the very same route. In order to see significant time savings, these slow bottlenecks should be upgraded. For example, a few 15mph curves and turnouts (Newark, Niles Jct, Coliseum) upgraded to 40mph would yield much greater time savings than their upgrading of 60mph curves to 75mph.

thetuck Wrote:
-------------------------------------------------------
> Increasing passenger speeds won't improve much
> unless freight speeds also increase. Otherwise
> you'd only create more conflicts with freight
> traffic.

That's not necessarily true.


> I agree there are many locations where
> passengers speeds could be higher, but most would
> not save enough time to justify the cost.


That depends upon the situation. For example, a curve good for F60 should be good for P74. On tangent, F60 is good for P80. Crossing signal start changes would be the expense.

>
> As for the upgrades mentioned on the UP Coast Sub
> in Norcal, the state paid handsomely for these
> upgrades which actually only save a couple minutes
> of running time, hardly worth the investment.


Once again, that depends upon a systematic assessment.

> Meanwhile very slow curves and turnouts still
> remain on the very same route. In order to see
> significant time savings, these slow bottlenecks
> should be upgraded. For example, a few 15mph
> curves and turnouts (Newark, Niles Jct, Coliseum)
> upgraded to 40mph would yield much greater time
> savings than their upgrading of 60mph curves to
> 75mph.

Absolutely yes. The slowest pieces are the first place to look.

TAW
 

SeaboardMan Wrote:
-------------------------------------------------------
> Superelevation isn't cheap.  Not just to install
> but maintain as freights are much heavier than in
> the old days.

That needs a systematic approach. I did an analysis for a Class 1 that found that if they reduced the superelevation for passenger trains, it affected the freight trains. It turned out that the thing to do was...nothing.

TAW

There are three ways to address make up time.

• The suggested method of raising speed but not changing schedules,
• Calculating running time at a speed lower than the maximum when developing the schedule (same thing but doesn't make physical changes and cost anything),
• Add recovery time after calculating the running time. Currently that is generally eight percent of running and schedule dwell for the whole run. Some folks call this padding. This is the current method of addressing the situation.

Under current railroad operating practices, I would not advise spending money to increase speed to increase reliability because it will be money shot into a vacuum, just as adding time to the schedule doesn't work.The real problem must be fixed, not mitigated.

TAW

while dispatching a busy single tracked ctc railroad,  amtrak at 79 mph "pounced" on slower 40-50 mph manifests.  it was very obvious watching it unfold on my screen.  i don't think increasing amtrak's speed would have done much good.

i think spending the money for 2 bi-directional main tracks in the same ctc territory with crossovers approx. every 10 miles would have provided more benefits to amtrak than raising any speed limits on an already "congested" single track railroad.........

66

I would still argue that passenger train speeds and freight speeds should not be the same as TAW pointed out. On most lines higher speeds are able to be handled for passenger trains than freight trains. Correcting these types of imbalances should not cost anything. For example, on the class 5 U.P. East Valley line north of Sacramento, the speed limit is 70 for freights and 70 for passenger trains. Technically class 5 allows 90 mph for passenger trains. Increasing the speed limit for passenger trains somewhat without any significant cost would not be unreasonable. 

I would also argue here that fast Amtrak bearing down on a 40-50 mph freight could eventually pass it and help the freight railroads. In stead of plodding along, it would quickly pass the next freights in the siding allowing them to proceed quicker onto the railroad. Taking 20 minutes instead of 30 minutes to get to the next siding cleans up the railroad for another train to use. 

A good place to consider this is UP's Sunset Area (former SP between El Paso and Yuma), where present passenger max is 79 and freight 70.
Much of this line has been double-tracked by UP and there are some long stretches of tangent track in AZ.
Another good candidate wouid be BNSF's former Santa Fe Bakersfield and Stockton subs. between Bakersfield and Port Chicago, CA.  This is also 79/70
track and had been good for 90 before Santa Fe removed ATS some time in the '60s.  This is the route of the San Joaquins.
Andy

goneon66 Wrote:
-------------------------------------------------------

> i think spending the money for 2 bi-directional
> main tracks in the same ctc territory with
> crossovers approx. every 10 miles would have
> provided more benefits to amtrak than raising any
> speed limits on an already "congested" single
> track railroad.........

Right. Look at the configuration of busy sigle track lines before CTC made many sidings "unnecessary." Many stations had two sidings so that trains could simultanously meet and be passed. CTC with one siding per station makes the execution of a meet and pass that should have occurred at the same place spread out over many miles and many, minutes.

Back before CTC increased capacity by ripping out track, double track railroads had two main tracks and either center sidings or sidings in both directions for the same reason. Santa Fe even had two main tracks and sidings in both directions in 261 (two track, no current of traffic) territory.

Two main track CTC isn't any good for overtaking when there is always a train coming the other way, regardless of what CTC salesmen said.

TAW

increasing amtrak's speeds from 79 mph to 90 mph on a single track ctc railroad would have meant "dumping" freights in the clear at least 1 siding before i normally would have to ensure no delays to amtrak.

overall, i don't think this would benefit the freight railroads.

i'd be interested to hear how other dispatchers and or current operating employees view this............

66





 

TAW Wrote:
-------------------------------------------------------

DELETED

> Two main track CTC isn't any good for overtaking
> when there is always a train coming the other way,
> regardless of what CTC salesmen said.
>
> TAW

from what i remember in 90 mph territory, i think that depends on how far apart the crossovers are placed...........

66

goneon66 Wrote:
-------------------------------------------------------

>
> from what i remember in 90 mph territory, i think
> that depends on how far apart the crossovers are
> placed...........
>


Right. In the first place, the CTC salesmen and signal engineers of the time placed sidings and crossovers according to distance instead of running time. Running time is the essential element of schedule and capacity, not distance.

Then, there is the amount of traffic, the combination of speeds, the frequency of trains of each speed. The standard method is pour trains (randomly or with a wide range of initial station leaving times at least) into a simulation and see what configuration generates an "acceptable" delay rratio. I've always used a trainsheet or traffic diagram (stringline) on paper to develop a solution, then tested with simulation.

TAW