State of Commuter Rail System

[dbperry]

At the 6/13/2016 MBTA Fiscal Management & Control Board Meeting, a presentation on the "State of the Commuter Rail System" was given.

Great stuff in here, especially the track diagrams on pages 32-33 showing dispatch responsibilities. Lots of other good data.
http://www.mbta.com/uploadedfiles/About ... System.pdf" onclick="window.open(this.href);return false;

Steve Koczela first pointed out the presentation, and had a brief but nice analysis of some of the data on his twitter feed:
https://twitter.com/skoczela/status/744949741306449920" onclick="window.open(this.href);return false;

He also very correctly points us all back to his analysis of MBTA Commuter Ridership data, which should be required reading for anyone with any even slight interest in the MBTA:
http://commonwealthmagazine.org/transpo ... nt-add-up/" onclick="window.open(this.href);return false;

[BandA]

Page 35 shows the Mean Moves Between Failure - The single level coaches and control cars are more reliable than the bi-levels despite being many years older. Perhaps due to the Rotems.

[Komarovsky]

Page 35 shows the Mean Moves Between Failure - The single level coaches and control cars are more reliable than the bi-levels despite being many years older. Perhaps due to the Rotems.

I noticed that as well, it is a bit disheartening. The chart of HSP vs everything else is quite interesting as well.

[BandA]

Need comparable MMBF from other systems.

[octr202]

I love how optimistic they are on the Haverhill Line. That's a lot more double track that is usable on that schematic!

[emannths]

Disappointed to see all the stats on a per-trip train basis instead of a per-passenger basis. A 30 minute delay on Sunday morning is a lot different than a 30 minute delay at 8am on Monday.

Of course, if you can't figure out how many people are on the train, it's hard to do the math.

[ohalloranchris]

The report also mentions "Remanufacture 10 Locomotives currently stored." Which locos are these? Based upon the list on NE Transit, could be later model F40's or screamers, although I would doubt the latter.

[Trinnau]

Disappointed to see all the stats on a per-trip train basis instead of a per-passenger basis. A 30 minute delay on Sunday morning is a lot different than a 30 minute delay at 8am on Monday.

Of course, if you can't figure out how many people are on the train, it's hard to do the math.

Industry standard is by trip, take a look at any other commuter rail network in the nation. Everything is geared to the trip measurement. We may see a shift over then next 30 years as technology changes, then you can even get into delay per passenger very precisely - which requires measurement at each station. For example an outbound rush-hour train that breaks down between its last two stations may have delivered 85-90% of its passengers on time, but the train is measured at end point so it's still late. So it's an imperfect science no matter what until the entire fleet is updated with reliable GPS and passenger counters.

Best breakdown right now is on the dashboard which shows peak performance assuming the vast majority of riders are in the peak periods.

I love how optimistic they are on the Haverhill Line. That's a lot more double track that is usable on that schematic!

If you're talking about slide 32, there are dashed lines there which appear to represent future work - you have to zoom in to see them. The schematic appears accurate for either before they retired Frye or after Vale comes online, but not the current condition.

[emannths]

Industry standard is by trip, take a look at any other commuter rail network in the nation. Everything is geared to the trip measurement.

There should at least be a breakdown of peak vs off-peak OTP. If you wanted to get fancy, you could look at the peak load on each trip (listed on dperry's turn sheet) and weight that by per-station boardings (slide 14 in the deck).

This is on my mind now because they've done a terrible job confining Worcester-line work to off-peak times. The stats already look bad, but the customer experience has been even worse. It was also an issue during the post-storm spring of 2015, when OTP stats improved faster than rush hour service.

[octr202]

I love how optimistic they are on the Haverhill Line. That's a lot more double track that is usable on that schematic!

If you're talking about slide 32, there are dashed lines there which appear to represent future work - you have to zoom in to see them. The schematic appears accurate for either before they retired Frye or after Vale comes online, but not the current condition.

Yes, I'm having some fun with that. They in fact do seem to indicate a very hopeful expectation that the first segment will be on line shortly. But see my signature line for my cynical thoughts on how long we've been waiting.

[Trinnau]

Industry standard is by trip, take a look at any other commuter rail network in the nation. Everything is geared to the trip measurement.

There should at least be a breakdown of peak vs off-peak OTP. If you wanted to get fancy, you could look at the peak load on each trip (listed on dperry's turn sheet) and weight that by per-station boardings (slide 14 in the deck).

This is on my mind now because they've done a terrible job confining Worcester-line work to off-peak times. The stats already look bad, but the customer experience has been even worse. It was also an issue during the post-storm spring of 2015, when OTP stats improved faster than rush hour service.

Looking at Peak vs. Off-Peak OTP gets a little tricky, simply because there are fewer peak trains. Most Lines have about 1/3rd of their weekday schedule as peak and the rest off peak, so the denominator is much larger when comparing off-peak service. There is also an inherent risk with more potential for failure during the rush hour. All available equipment is in service running at a greater frequency resulting in much greater risk of delays reverberating through the schedule. The same failure on a Wednesday morning might severely impact service but on a Saturday might only impact 1 or 2 trains, for example the car on the tracks on the Worcester Line yesterday. But of course all those trains are running because that's when the people use them the most, so it's a bit of a double-edged sword in that way. People want the trains and they should perform, but as a result there is also a higher chance for failure and a higher chance for additional delays.

As I said above, MBTA's performance dashboard splits up the peak and off-peak service. I think you'll be surprised to find the peak performance is always going to be worse than off-peak performance for the reasons I mentioned. For example, 1 late peak train on the Worcester Line gives a peak OTP of about 95% (20 of 21 trains) where 1 late off-peak train gives an off-peak OTP of about 97% (34 of 35 trains).

[bierhere]

They consider West Natick assessable with its crumbling closed off mini-high. I haven't seen them use the portable lifts they currently have on the platform there, like the Green line stations. Are there any plans to fix the mini-high in West Natick?

[octr202]

Makes me wonder how many mini-highs and other portions of boarding platforms are out of service. Only Andover appears to be noted in T Alerts. It's mini-high was closed off several months ago, edge bolted up. I guess we wait until there's funding found to repair it, perhaps in the next fiscal year.