Overhauling Amtrak

[Alon]

The flip side is that Chicago-Detroit has so many slowdowns, of which the worst is in Chicago itself, that the top speed isn't that important. At the upper limit of what you can do when you have no slow zones, Tel Aviv-Haifa express trains average about 120 km/h with diesel locos that top at 140 km/h. That line is straight, which helps, but it also has no terminal slowdowns, since Tel Aviv's stations were reconfigured as through-stations in 1996.

[Dr Nevergold]

Originally Posted by Evergrey

Uhhh... Pittsburgh is way closer to the allegedly "isolated" major cities of the East Coast than Chicago. DC is particularly close - closer than Detroit, a city most Pittsburghers have no reason to travel to. It has nothing to do with being "hell bent". People are making rational travel decisions... Pittsburghers are more intertwined with the East Coast than with the Midwest.

Look at your own situation, JonBoy... do you live in Cleveland or Detroit? No. You're a Pittsburgher living in suburban Philly. Pittsburgh's primary nodes of population and economic interchange are DC/Philly/NYC. This is driven by generations of labor markets, migration patterns, business development, universities, etc. Cleveland might be close... but there isn't a whole lot going on between PGH-CLE to prioritize such a route above eastern connections... and there's much less connectivity between PGH and more distant Midwestern cities. Toledo? South Bend? Are you kidding me? Pittsburgh's orientation is to the east.

One of the "obvious reasons" for this orientation is detailed in this 2010 study by Pitt's University Center for Social and Urban Research:

That's a pretty astute analysis about Pittsburgh's travel corridors. Pgh has always been linked closer to the east coast than the midwest, probably because its in a northeastern state even if its on the western edge. Most everyone I know of considers Pittsburgh a northeastern city, even if it isn't "east coast" per se.

Cleveland and Pittsburgh are noticeably different cities, but they are very close. Only 120 miles between the downtowns, so I bet there's enough traffic to warrant it if they build it.

I'd like to see a high speed northeastern triangle: the NE corridor from DC to Boston, then a high speed rail link out west to Pittsburgh, Cleveland, then up to Buffalo (connecting with Canada and the Toronto area with transfers) then east toward Boston, and maybe a high speed service between NYC and Montreal connecting at Albany right in the middle.

New York is a mountainous state, but the corridor that the train could travel is along the 'coastal plain' of the Great Lakes. Its relatively flat between the shores of Lake Erie and Lake Ontario and several miles inland, south of Buffalo and Rochester. Wouldn't be too expensive to cut through the northeastern rugged terrain in upstate thanks to the land between the lakes and the mountains.

[Nexis4Jersey]

Originally Posted by Alon

I question whether NY-Buffalo will really be built before Philly-Pittsburgh. On the federal level, nobody's yet committed to even looking at HSR on these corridors, and a cost-benefit calculation should show that they are close in both costs and benefits. On the state level, neither state is interested in HSR, and the Cuomo administration just explicitly rejected studying HSR further on the grounds that the state can't afford it. (In Cuomo-land, the economic conditions of 2012 should dictate what spending the state should take on in the 2020s.)

When did Cuomo cancel the study?

[ardecila]

Originally Posted by Alon

The flip side is that Chicago-Detroit has so many slowdowns, of which the worst is in Chicago itself, that the top speed isn't that important. At the upper limit of what you can do when you have no slow zones, Tel Aviv-Haifa express trains average about 120 km/h with diesel locos that top at 140 km/h. That line is straight, which helps, but it also has no terminal slowdowns, since Tel Aviv's stations were reconfigured as through-stations in 1996.

Correct; the route between Chicago and Porter, IN runs over Norfolk Southern trackage and passes some busy yards.

Between CREATE and the IDOT rail projects, many of these slow zones will be eliminated. MDOT also has some money to reduce slow zones on the Detroit end, too. The general curviness of the line can't really be helped, though. It's not as bad as the Shore Line throuch CT/RI.

[Standpoor]

^
The Detroit part is awful. I very rarely ride past Ann Arbor but my past couple of rides have taken me to Royal Oak and the Chicago/Northwest Indiana sections were breezes compared to the sections east of Ypsilanti. We never broke forty between Ypsi and Dearborn, and once in Detroit the track is abysmal, rough and I presumed jointed. Couple this with 15 mph and 8 mph maneuvers and I was ready to get off by the time we reached Royal Oak. Track improvements and rehabbing the connection on Detroit's west side (forgive me I don't know what it is called), will really help with on time performance, especially on westbound trains.

With that being said, I am glad that they are starting with the section between Kalamazoo and Battle Creek since it is so slow. The scheduled average speed is 35 mph. So immediately after sustained 100 mph operation, the train slows down and crawls the 22 miles between the stations.

Finally, to bring it back around to overhauling Amtrak, one thing I wish Amtrak was better at doing is telling you where the trains are on their route. A few weeks ago, I was in St. Joseph, MI and wanted to take the Wolverine back to Chicago. The Wolverine low bucket price is about 22 percent lower per mile than the Pere Marquette, so I opt to take the Wolverine out of New Buffalo. Because of the Wolverine's tradition of tardiness, I waited until the last minute to buy my tickets. The train was 25 minutes late into Battle Creek and thinking that 25 minutes was not bad, I purchased my tickets. Print at home tickets are a game changer at unstaffed corridor stops.

I arrived at New Buffalo about 15 minutes after scheduled departing time knowing that the train still had not made the stop in Niles(27 miles away) thanks to Amtrak app. However, the train had equipment problems around Niles and did not get to New Buffalo until an hour and a half later. The problem was the Amtrak website kept updating the estimated time of arrival by 5-10 minute increments. It would say estimate arrival 3 pm and once 3 pm rolled around would update to 3:10 and so forth for an hour and a half. Which means I just sat there the whole time. If Amtrak had some sort of gps tracking system, I could have grabbed a bite to eat or drink and then once the train started moving again, I could have made my way back. It was very frustrating not knowing where the train was. On short corridor trips, knowing when precisely a train will arrive is very important.

[electricron]

Originally Posted by hammersklavier

Recall, electricron, that most nations that develop true HSR corridors do so atop a large-scale medium-speed network. This is true in Italy, Germany, Sweden, China, S. Korea, and Great Britain, and even in the earliest adopters--Japan and France. About the only nation where this paradigm does not hold is Spain, because (as Alon--Levy, I assume?--pointed out, Spanish HSR was built simultaneously with extensive improvements to its medium-speed service).

Also, may I remind you that for super-110 operation, electrification is a requirement? This ups the costs quite a lot; it is unlikely to happen so long as the freight railroads rely on diesel service (because diesel power will remain the primary prime mover in this country).

What I think is most likely to happen is that we'll continue to see upgrades to 110 mph service, and somebody (or somebodies) developing a clone/upgrade of the IC 125 which will be hailed as "a revolution in rail transportation!" , with post-110mph service mainly limited to the Northeast and California markets until a large enough network, a large enough proven market, exists in the Midwest for it to independently electrify its five primary routes (Chicago-St. Louis, Chicago-Minneapolis, Chicago-Detroit, Chicago-Cleveland, and Chicago-Indianapolis-Cincinnati). That would not happen until ca. 2030 at the earliest.

But, of course, massive changes can happen. What if the price of gas plateaus at $5/gal? The major freight railroads would have intense pressure to electrify.....

Rail.corridors with 125 mph operations are usually quad track if the lines are used for other purposes. The northwestern corridor in Great Britain is quad rack where congested, as is America's northeast corridor. Slower trains running on the outside tracks with faster trains running in the middle tracks. Overtaking one another is safer than when faster and slower trains are not running on the same track and avoid being switched onto another track. Unless the Midwest lays quad track rail corridors, you're not going to see 125 mph operations just due to the large differences in train speeds. Additionally, FRA regulations frowns on faster than 110 mph operations on tracks with at grade crossings. Again, the Midwestern states are never going to eliminate at grade crossings along the entire corridors. A few here, a few there, will be eliminated, but no where near enough to affect max train speeds.

I wrote earlier increasing max speeds to what the owners of the corridors will allow. Both UP and BNSF have stated 90 mph as that limit on shared tracks. So, even 110 mph is unlikely on freight owned tracks (the sole exception being Chicago to St. Louis UP owned corridor). 110 mph speeds will only be reached on public owned corridors, ie the Michigan owned line to Detroit.

FEC All Aboard Florida trains will reach 110 mph where their corridor isn't congested, and 125 mph where only passenger trains will be running. That's the best you"ll ever see on a freight owned corridor - and technically where they will be going 125 mph, that section of the corridor will be owned by Florida - FEC will only be leasing it.

[Alon]

On the NEC, one straightforward thing Amtrak and the commuter railroads should do is let people know in advance which train is departing from which track. In Germany, Japan, Switzerland, and others, when you buy a train ticket it includes the track number, because part of the timetable is about how trains go through city stations, and when there are conflicts, station throat improvements are built with specific schedules in mind. At some major US stations, there is a natural division into tracks that makes it very easy to adopt this practice, including 30th Street (Amtrak gets 10 tracks all for itself), Grand Central (a gazillion tracks), and South Station (4 lines - Worcester, Providence/NEC, Fairmount, Old Colony - side by side). Penn Station's a bit harder, but it has 21 tracks, which is more than enough for the traffic it has; in Tokyo there are 4-track train stations with considerably more traffic and even 2-track stations.

[LMich]

Originally Posted by Standpoor

Track improvements and rehabbing the connection on Detroit's west side (forgive me I don't know what it is called), will really help with on time performance, especially on westbound trains.

West Detroit Connection Track. Construction starts at any time, now, as "early 2013" is the scheduled. The project will seperate to a point freight and passenger manuevers at the junction, and includes new passenger tracks and a bridge. The crazy manuever at West Detroit, currently has the train passing the junction, then having to back up, and then pull unto the line to New Center.

As for Milwaukee Junction - the junction Amtrak has to navigate to get up to the northern burbs - there isn't much that can be done, there. It's just how the metro is layed out.

[Nexis4Jersey]

Originally Posted by Alon

On the NEC, one straightforward thing Amtrak and the commuter railroads should do is let people know in advance which train is departing from which track. In Germany, Japan, Switzerland, and others, when you buy a train ticket it includes the track number, because part of the timetable is about how trains go through city stations, and when there are conflicts, station throat improvements are built with specific schedules in mind. At some major US stations, there is a natural division into tracks that makes it very easy to adopt this practice, including 30th Street (Amtrak gets 10 tracks all for itself), Grand Central (a gazillion tracks), and South Station (4 lines - Worcester, Providence/NEC, Fairmount, Old Colony - side by side). Penn Station's a bit harder, but it has 21 tracks, which is more than enough for the traffic it has; in Tokyo there are 4-track train stations with considerably more traffic and even 2-track stations.

They do , at most stations. Sometimes a few days in advanced at the busy stations like Newark Penn , Baltimore Penn , Stamford...etc , the exception to this is New York Penn Station which is 25mins before departure. Amtrak shares with NJT at 30th Street... Not all 21 tracks in New York are used , about 7 are storage tracks for emergency equipment , or layover trains for Lake Shore Limited or Empire Service.... Then the LIRR gets a few tracks and Amtrak gets a few tracks and NJT gets a few tracks....its the same with Grand Central , only 15 of those tracks are used most of the day , during rush hour about 30 tracks are used but the rest of the tracks are used for storage , its mostly a yard. They also have equipment down there and Freight cars....

[ardecila]

Originally Posted by LMich

As for Milwaukee Junction - the junction Amtrak has to navigate to get up to the northern burbs - there isn't much that can be done, there. It's just how the metro is layed out.

I'm not an expert on the rail network of Detroit, but between Brush St and Milwaukee Junction, Amtrak must cross Norfolk Southern, then cross back. There appears to be space in the ROW to shuffle some tracks and eliminate the weave. It's a bigger project, on par with what Chicago is doing at Grand Crossing. It's definitely achievable, though - probably $50-60M?

[Wizened Variations]

Originally Posted by electricron

Rail.corridors with 125 mph operations are usually quad track if the lines are used for other purposes. The northwestern corridor in Great Britain is quad rack where congested, as is America's northeast corridor. Slower trains running on the outside tracks with faster trains running in the middle tracks. Overtaking one another is safer than when faster and slower trains are not running on the same track and avoid being switched onto another track. Unless the Midwest lays quad track rail corridors, you're not going to see 125 mph operations just due to the large differences in train speeds. Additionally, FRA regulations frowns on faster than 110 mph operations on tracks with at grade crossings. Again, the Midwestern states are never going to eliminate at grade crossings along the entire corridors. A few here, a few there, will be eliminated, but no where near enough to affect max train speeds.

I wrote earlier increasing max speeds to what the owners of the corridors will allow. Both UP and BNSF have stated 90 mph as that limit on shared tracks. So, even 110 mph is unlikely on freight owned tracks (the sole exception being Chicago to St. Louis UP owned corridor). 110 mph speeds will only be reached on public owned corridors, ie the Michigan owned line to Detroit.

FEC All Aboard Florida trains will reach 110 mph where their corridor isn't congested, and 125 mph where only passenger trains will be running. That's the best you"ll ever see on a freight owned corridor - and technically where they will be going 125 mph, that section of the corridor will be owned by Florida - FEC will only be leasing it.

Through the years, I have been confounded at the public's perception of peak velocity versus average velocity when taken between 2 or more points along a dedicated route.

A very straight forward example is a light rail system with a maximum velocity of 60 mph, and takes 20 minutes to go 10 miles, including stops. That's a 30 mph average.

I apologist for being simplistic, but this illustrates the most important step in improving our passenger rail system: the need to increase the average speed through urban areas, including boarding and deboarding at stops, station approaches (through or stub), etc. This is the first area where progress can be made to get the fastest results.

Let's say the freight railroads put a maximum speed of 90 mph even through urban areas which might then be maintained (i.e., 90 mph for a 100 mile stretch for example). If approaches to stations from 90 mph to 0 could be made in say 2 miles on either side of a stop, and, the stop could be made in 2 minutes or less (very routine time length), then for 4 miles out of 104, the average speed would be 15 mph, and the 104 miles traveled in 73 milnutes for an average speed of 85 mph*

I have felt for a long time that what needs to be done is to increase the average speed on all Amtrak routes 1 mph per year.

Amtrak has to take numerous small steps to repair damage caused by property development, tracks being ripped up or downgraded, etc., over almost 60 years.

Amtrak (and the US in general) needs the humble approach. The public rail net work needs to both add new 79 mph routes, and, relentlessly remove bottlenecks to increase average speed. **






*(66.7 minutes for the 100 miles and 6 minutes for the 4 miles in a total time of 72.7 say 73 minutes or 1.22 hours which averages out to 85 mph over that 104 miles).

**in 20 years the change would be very evident.

[nei]

Originally Posted by Alon

On the NEC, one straightforward thing Amtrak and the commuter railroads should do is let people know in advance which train is departing from which track.

Eh. I get the feeling whoever is running Penn Station doesn't know what track the train is coming in until it arrives. Doesn't make a huge difference, the trains usually sit there for at least 5 minutes.

[Troyeth]

Originally Posted by nei

Eh. I get the feeling whoever is running Penn Station doesn't know what track the train is coming in until it arrives. Doesn't make a huge difference, the trains usually sit there for at least 5 minutes.

Those minutes accumulate and can become particularly disruptive of efficient, high-capacity, high-performance passenger operations.

[Nexis4Jersey]

Originally Posted by nei

Eh. I get the feeling whoever is running Penn Station doesn't know what track the train is coming in until it arrives. Doesn't make a huge difference, the trains usually sit there for at least 5 minutes.

This is the typical madness at Penn station all trains are annouced 5-25mins before departure causing a stampede....

• Video Link

• Video Link

[Alon]

Don't forget another problem of Penn: there aren't a lot of official access points to most platforms. Amtrak funnels passengers through just one access point on the upper concourse, where employees check their tickets (in addition to an on-board check). This slows boarding by several minutes, minutes that are more comparable to lining up for security at airports than to sitting in a meter-pitch seat on a train. There are other access points on both concourses, but these are unannounced, and you need to know what you're doing to even know that there's a train at the platform you're going down to.

[TTU Arch]

A new report has come out about Amtrack's ridership numbers from the Brookings institute. http://www.brookings.edu/research/re...-puentes-tomer

Its a mixed bag but generally ridership numbers have gone up. I have to say being in Texas I had to do a double take at the DFW numbers. While DFW is not one of the 10 metro's they cite as being a major contributer to ridership, the overall increase from 97 to 2012 is substantial (482.9% up from 34K to more than 200K). Something to not be ignored and shows growing support of passenger rail at least in North and Central Texas. Houston had a growth of a little over 4,000 riders in the same period to a total of 20K plus riders, San Antonio 60% to 70K, and Austin 383% to about 54K. I'm a supporter of HSR, and now understand why Amtrack in Texas is concentrated more on connecting DFW with San Antonio / Austin with HSR. I always thought Houston / DFW was the first natural connection, but looking at the ridership numbers puts a new perspective on it. Still think Houston and DFW need to be connected as being the two primary metro's in the state.

http://transportationblog.dallasnews...il-plans.html/

[aquablue]

Originally Posted by Alon

Don't forget another problem of Penn: there aren't a lot of official access points to most platforms. Amtrak funnels passengers through just one access point on the upper concourse, where employees check their tickets (in addition to an on-board check). This slows boarding by several minutes, minutes that are more comparable to lining up for security at airports than to sitting in a meter-pitch seat on a train. There are other access points on both concourses, but these are unannounced, and you need to know what you're doing to even know that there's a train at the platform you're going down to.

In Europe they don't have the gate system, they just let people walk on the platform without being funneled through a gate. This seems to be a better way to go and speeds up boarding. However, these stations are usually far more roomy with bigger platforms.

[Alon]

On narrower platforms, multiple access points are even more important. The traffic on intercity trains isn't enough to overwhelm even the Penn platforms, as long as people can disperse into the concourse easily. The problem is when people from many different cars come together in search of one escalator.

[PhillyRising]

Got a chuckle watching the videos of the crowds at Penn Station. For NJ Transit trains...everyone just stands around trying to figure out what track their train may be leaving from and when they announce it it's this tsunami of humanity pushing towards the two little doors to get down to the train. It's just madness.

[GreatTallNorth2]

Originally Posted by aquablue

In Europe they don't have the gate system, they just let people walk on the platform without being funneled through a gate. This seems to be a better way to go and speeds up boarding. However, these stations are usually far more roomy with bigger platforms.

The European system is much better in this regard. I live in the UK now and took the train every day to work. Trains wait at the station for about 2 minutes maximum. This is enough time for people to get off the train and for passengers to board. They have gates (like subways) where it checks your tickets.