Market-Frankford Service Issues

I think its time to open a discussing regarding the current operation on the MFL. The line has tons of issues right now, and considering its the most used line on the whole system I think it deserves more attention from SEPTA.

I've been noticing a pattern of the line becoming more crowded over the past few months. Even trains at 10:30am are standing room only at times. Rush hour is even worse. I've heard a rumor that SEPTA is investigating the possibility of running express limited service, such as local stops in Center City, then express to FTC with stops at Girard, Allegheny, Margaret Orthodox and FTC.

Something definitely needs to be done. There is a new 4 or 5 story parking garage that will be opening soon at Frankford. If that garage fills to capacity every day, trains will be rolling out of FTC stuffed. Would their be much point running trains filled to capacity into CC and making local stops? Would seem to make much more sense making those trains express, if that's even possible.

There is also a huge issue of deterioration of both the cars and infrastructure. Track speeds have been reduced between Girard and 2nd Street on what seems to be an indefinite period. It also seems that either Philadelphians are bigger slobs now than they were 5 years ago, or SEPTA isn't spending as much time cleaning the cars (or perhaps they're using them as a new petri dish for a bacterial experiment).

Thoughts from everyone....

jfrey40535 wrote:I think its time to open a discussing regarding the current operation on the MFL. The line has tons of issues right now, and considering its the most used line on the whole system I think it deserves more attention from SEPTA.

I've been noticing a pattern of the line becoming more crowded over the past few months. Even trains at 10:30am are standing room only at times. Rush hour is even worse. I've heard a rumor that SEPTA is investigating the possibility of running express limited service, such as local stops in Center City, then express to FTC with stops at Girard, Allegheny, Margaret Orthodox and FTC.
....

I hope they do run that express train espscially atleast starting at 6:30 Am...that 7AM A/B service just does not cut it for me sometimes.
Because the BSL has express service starting @ 6:15AM....though it does have more tracks...but half the time the MFL Eastbound is like two mins. late after its scheduled time @ 30th st.

jfrey40535 wrote:I've heard a rumor that SEPTA is investigating the possibility of running express limited service, such as local stops in Center City, then express to FTC with stops at Girard, Allegheny, Margaret Orthodox and FTC.

What will happen when an express finds itself one station behind a preceding local? Unless someone can find a magical way to add more tracks the MFL will always be constrained by the slowest train running. The A-B station arrangement, at least before a lot of the stations were changed to all-stops, was about the best that could be managed and it's pretty cumbersome.

There's a passing siding at Spring Garden. I'm not sure how useful it would be. Maybe they could build another passing siding further north.

Lucius Kwok wrote:There's a passing siding at Spring Garden.

Great diagram, Lucius, but that's a pocket siding for turning trains back, not a passing siding. And as you suspect, it's not a particularly useful location for facilitating express service to/from Frankford.

Practically speaking, an A/B operation is the most feasible way of speeding up service to Frankford (Vukan Vuchic's textbook has a good description of the options), but SEPTA has watered down the benefits by making more stations AB.

The way to figure out where sidings would be necessary would be to draw a stringline diagram of the line, which you can do well enough in Excel. Set up two rows for each station: when a train arrives and when it leaves, then fill down the columns with times (you'd want to ride a trip or two using a stopwatch to clock typical operation. Put distance (blocks or miles) in the first column. Then paste in more columns showing more trains departing at intervals of your desired headway. Then set up columns for your express trains, which will have to slow down, but not stop for the local stations. Once you have an hour or so of trains in the spreadsheet, pull up the chart command, choose an XY chart with lines and no point markers, make the category axis for your first series the times of the first train, and make the value axis distance. Copy the series, paste it in the graph again, and change the category axis to the times for the second train. Repeat until you have the full stringline. (e-mail me, and I'll send you a copy of the one I did for SEPTA's Schuylkill Valley plan--you can use the graphs off it)

Once you have the diagram, everywhere you see one line crossing another is a place you need to put a siding. Insert time rows and times at the siding so the local train holds for the express train, and then you'll see the location of the second siding move. Repeat for the next siding, etc. Then, by changing the times at the top of each column, you can model how the line flows at different headways or when one train gets late. You'll find that sometimes a small change in headway has a huge effect on whether the meets work.

This is a good exercise for all of us armchair engineers and dispatchers--a chance to see a bit of the complexity in designing and scheduling a railroad. With the exception of SEPTA and SVM, this is normally a very important part of planning a railroad. Important enough that when NJ Transit planned the Camden-Trenton line, they actually did their own stringline as a check on the ones the consultants drew up.

BTW, my assumption would be that trains going the opposite of the peak direction would all run local, so you would only need to build one set of sidings between the two mains. If the MFSE is getting a lot of reverse commuters (don't know if it is or not), that assumption may have to be revisited.

Matthew, thanks for the tips. I will be doing some work Excel soon.

I have a copy of Edward Tufte's books and one has some stringline charts, so you don't need to email a chart.

One example of a two-track railroad with express service that passes locals is the Caltrain Baby Bullet service. They built couple passing sidings to do that. Now, that is a commuter rail service so it has only five trains per hour in each direction.

Lucius Kwok wrote:One example of a two-track railroad with express service that passes locals is the Caltrain Baby Bullet service. They built couple passing sidings to do that. Now, that is a commuter rail service so it has only five trains per hour in each direction.

Yep. And setting up the meets there is a ton easier than setting them up when you have headways of 3 minutes or less. Another thing I didn't mention is that when you have such close headways, the signal configuration is going to have a big effect. When the second train leaves the siding, it has to wait for the block ahead to clear. Therefore, you'd want the siding to be located just before a local stop, so the local could do its station work while waiting for the express to clear out. Bottom line is that while a back of the envelope calculation like what we're talking about can show that a proposal is not technically feasible, I wouldn't rely on it to show that it is.

It's interesting that the SEPTA Route 100 (NHSL/P&W) manages to have three types of service (Bryn Mawr local, Hughes Park express, Norristown limited) on a two-track railway. It seems they're able to run trains only one minute apart, and have the local go right after the express, and do this with 12 trains per hour in each direction.

It would be possible to have an express in the Frankford El without passing maneuvers if the trains are far enough apart. In practice, this would probably break down rather quickly.

The alternative is skip-stop service. But if too many people have to change trains because of skip-stop, it negates the speed advantage.

If the Roosevelt Blvd subway were built, it would have taken some of the demand off the El.

Lucius Kwok wrote:It's interesting that the SEPTA Route 100 (NHSL/P&W) manages to have three types of service (Bryn Mawr local, Hughes Park express, Norristown limited) on a two-track railway. It seems they're able to run trains only one minute apart, and have the local go right after the express, and do this with 12 trains per hour in each direction.

12 trains per hour (15 minutes between fleets) is a lot different from 20-24 trains per hour. Another thing is that those locals are turnbacks--on the El, everything has to go to Frankford (which is where the greatest ridership is).

It would be possible to have an express in the Frankford El without passing maneuvers if the trains are far enough apart.

That's an enormous if.

In practice, this would probably break down rather quickly.

And the tighter the headways, the less margin you have for error.

The alternative is skip-stop service. But if too many people have to change trains because of skip-stop, it negates the speed advantage.

Skip-stop works because most passengers are going to the CBD.

If the Roosevelt Blvd subway were built, it would have taken some of the demand off the El.

Yep. Got 3 or 4 billion to spare?

Using the Capital Cost report from the Roosevelt Blvd Corridor web site (which is no longer up), the cost for the 12.3 mile BSL extension plus a 0.8 mile extension of the MFL as a subway is $3.4 billion.

But if we just build the BSL extension to Cottman Avenue, I figure the cost is $1.5 billion.

But I have the impression that the most vocal Northeast Philly residents don't want a subway up there, since it would bring "undesirable elements" and "export crime" from North Philly to their neighborhoods, so the political reality is that it's not going forward. I guess they'll have to live with what they have.

Matthew Mitchell wrote:

Lucius Kwok wrote:One example of a two-track railroad with express service that passes locals is the Caltrain Baby Bullet service. They built couple passing sidings to do that. Now, that is a commuter rail service so it has only five trains per hour in each direction.

Yep. And setting up the meets there is a ton easier than setting them up when you have headways of 3 minutes or less. Another thing I didn't mention is that when you have such close headways, the signal configuration is going to have a big effect. When the second train leaves the siding, it has to wait for the block ahead to clear. Therefore, you'd want the siding to be located just before a local stop, so the local could do its station work while waiting for the express to clear out. Bottom line is that while a back of the envelope calculation like what we're talking about can show that a proposal is not technically feasible, I wouldn't rely on it to show that it is.

In fact, the best configuration is to have the station within the limits of the siding, so the expresses can bypass the locals sitting in the station. Of course, that adds a whole new level of complexity and cost - moving stations to be situated at passing sidings.

I think you're right - Vuchic's textbook explains the options well. You're also right in that SEPTA misses the boat by making too many stations AB.

But I have the impression that the most vocal Northeast Philly residents don't want a subway up there, since it would bring "undesirable elements" and "export crime" from North Philly to their neighborhoods, so the political reality is that it's not going forward. I guess they'll have to live with what they have.

And what effect do they think the Route 14/20/1 Bus routes have? Instead they are forgoing themselves a means to reach Center City without having to deal with the Blvd. And those people need to wake up as that part of the northeast is already in decline, and its not due to a non-existant subway. Our transit system needs to stop falling to its knees because of some old lady in a wheelchair who doesn't want a subway or is mad because the bus/trolley doesn't stop in front of her house anymore.

jfrey40535 wrote:Our transit system needs to stop falling to its knees because of some old lady in a wheelchair who doesn't want a subway or is mad because the bus/trolley doesn't stop in front of her house anymore.

It ain't the transit system falling: it's City Council and the brass at SEPTA. And the mayor too: by making transportation an even lower priority than it was in Rendell's administration, and Goode's before that, he's effectively defaulted all transportation decisions to the ten district City Council members.

Getting back to the original topic, I'm thinking that there is one traditional means to carry more people on a subway car: Remove seats. Have bench seating. You can fit more people standing than sitting. Sure, it is less comfortable, but it doesn't cost as much as building more tracks.