1) Worcester Line Working Group (recap)
2) Why are all AM trains late on this new schedule?
3) Results of the public comment process
1) Worcester Line Working Group (and how we got a new schedule)
First of all, it's important to remember that the schedule changes are only made twice a year: in May and November. This is because every time the schedule changes, the entire crew rotations for all lines need to be changed - which takes time and effort. I believe that there are contractual obligations for the crews to have a six month duration for their assignments, also. Most crewmembers do not stay on one line throughout their workday, so it's really only possible to change all the schedules for all the lines at the same time. There have been minor schedule changes published in other months, primarily to accommodate construction or to correct problems with published schedules.
With that in mind, here's a recap timeline of how we got to the new 5/22/2017 schedule:
October 2015: Lt. Governor Karyn Polito announces a new once daily roundtrip non-stop Worcester to/from Boston "bullet" train. It was soon renamed the "Heart to Hub." By the time this new concept was announced in October, it was too late to implement it in November 2015 - the schedule was already set and crew assignments were well underway.
January 2016: The MBTA publishes a draft schedule to be implemented in May 2016 that incorporates the new Heart to Hub service.
February 2016: Public hearings are held and a public comment period is implemented to gather feedback regarding the new draft schedule.
May 2016: The new schedule is implemented that incorporates the Heart to Hub train (along with some other (not minor) changes).
June 2016: Minor tweaks are announced to the May 2016 schedule. These tweaks increase the duration of certain trains, in recognition that the May 2016 schedule was too aggressive and not consistently achievable.
August 2016: I publish a proposal to modify the morning schedule. I eventually gather about 600 signatures on a change.org petition to implement the proposal.
September 2016: With the volume of public interest in the Framingham-Worcester line schedule increasing, as well as upcoming infrastructure changes for the line (Boston Landing and the double track through Beacon Park), MBTA General Manager Brian Shortsleeve decides to keep the current May 2016 schedule (with the July tweaks) in place and not make changes in November 2016. At the same time, he invents and convenes the "Worcester Line Working Group."
September - December 2016: The Worcester Line Working Group meets approximately bi-weekly to discuss potential changes for the schedule. The Worcester Line Working Group eventually included: Lt. Governor Karyn Polito, Rep. Alice Peisch (D - Wellesley), Rep. Carolyn Dykema (D - Southboro, Westboro), Rep. James O'Day (D - Worcester), Stuart Loosemore (Worcester Chamber of Commerce), Jessica Strunkin (495/Metrowest Partnership), Brian Shortsleeve (MBTA General Manager), Jody Ray (MBTA Assistant General Manager in charge of Commuter Rail Railroad Operations), two Deputy Directors of MBTA Railroad Operations, David Scorey (Keolis CEO/GM), the Keolis Manager of Operations Planning, a general public member / commuter from Wellesley, and me.
December 2016: A draft schedule for May 2017 is published.
January - March 2017: Public hearings are held and a public comment period is implemented to gather feedback regarding the new draft schedule.
March 2017: The Worcester Line Working Group reconvenes to review the public comments and finalize the May 2017 schedule. See item #3 below.
May 2017: The new schedule is implemented. Whew!
2) Why are all AM trains late on this new schedule? (AKA: It's not my fault. Really!)
As almost everyone has noticed, the AM on-time performance has been less than stellar throughout May and June. In fact, almost every train is late in the morning. Believe it or not, it isn't yet possible to blame the new 5/22/2017 schedule.
As fate would have it, the last stretch of the line that needed tie replacement and railbed resurfacing was in one of the most critical stretches for us: from Wellesley to Southboro on track 2, which is the inbound track for the morning rush hour. This project started in April 2017 and should be fully completed as of Friday 6/23/2017 (although I'm not sure when or if they will complete the tie replacement directly within Framingham station, which appears to have stalled).
As described in this blog post, tie replacement and railbed resurfacing requires speed restrictions on the disturbed area. The MBTA has implemented policies and procedures to minimize the length of the speed restriction, but it can't be eliminated.
These speed restrictions are directly responsible for almost all of the morning delays we have experienced in May and June. Obviously there have been significant isolated events such as equipment failure, but I'm talking about the daily 5-15 minute delays that have been persistent.
One of the recent questions I have heard is that once the construction work and speed restrictions moved west of Framingham, why were local trains that originated in Framingham (and therefore didn't have to encounter the speed restrictions) affected?
The reason for this is that the AM schedule is now "maxed out" and a delay to any train can easily cascade to following trains. What do I mean by "maxed out?"
First of all, it's important to understand the basics of automatic block signals. That sounds technically complicated, and I describe it in detail in the glossary, but it's a simple concept.
The line is broken up into blocks, with a signal at each end of each block. Very generally, the concept is that only one train can occupy each block at a time. So let's imagine a train line with 4 blocks: A, B, C, and D:
Block A: occupied by a train
signal between block A and block B: Red, to indicate that a train in block B shouldn't enter block A
Block B: empty
signal between block B and block C: Yellow, to indicate that a train in block C should go slow through block B and be prepared to stop at the red signal between A & B.
Block C: empty
signal between blocks C and block D: Green, since blocks C and B are empty.
Block D: the next train
(This is highly simplified and just here to illustrate the concept. It gets a little more complicated in real life, but this suffices for the purposes of our discussions here. In other words, railfans and railroad employees, don't fill up the comments below with how a red automatic signal doesn't prevent a train from passing it, etc. etc.).
Now consider that on the AM inbound schedule, local trains depart just after the express trains pass through Framingham (it would be somewhat silly to put a local train in front of an express!). Using the block concept above, it is obvious that the local train can't be scheduled to leave immediately after the express train - otherwise it would be stuck at a red light until the express got to the next block and then the local would still have a yellow signal (and slow speed) until the express got 2 blocks ahead.
On the Framingham-Worcester line, the block length (and therefore signal spacing), speed limits, and other considerations associated with the automatic block signal system mean that ideal minimum time between trains is 8 minutes. This interval is the most crucial element for the AM schedule.
Take a look at the intervals between express and local trains at Framingham in the morning on the new 5/22/2017 schedule: there are intervals of 8-9 minutes for each pair of a local train following an express train. Therefore, it is critical that the express train gets through Framingham on time - otherwise, almost any delay to an express train will delay the local train behind it.
Why is the schedule built with the bare minimum spacing between local and express trains at Framingham? The answer is at Yawkey, where the expresses meet the locals at the other end of their express runs. Taking a look at the interval between expresses and locals at Yawkey reveals intervals of 10-12 minutes - close to that minimum 8 minutes for block spacing. The locals need to leave Framingham early enough so that they can get to Yawkey early enough to not delay the following express train.
This is how the schedule is "maxed out" from 5:15 AM to 9 AM. There isn't any room on the schedule to insert another train - the local trains are just barely squeaking between the express trains. The local trains depart Framingham at the minimum interval AFTER an express, and arrive at Yawkey close to the minimum interval BEFORE the next express.
The good news is that we now have every possible train we can get, given the current infrastructure. This means the frequency of trains is maximized. The bad news is that delays to any one train can quickly cascade to every other train following it, and that is exactly what has been happening with the recent tie replacement project.
With an express train encountering a 1-2 mile speed restriction in Southboro as a result of the tie replacement project, it becomes 5-10 minutes late. That's enough to delay the departure of the local train from Framingham, and then the local train is 'chasing the signals' of the express train until the express can get far enough ahead of the local train. For example, if the local train departs Framingham 5 minutes late due to the late express train, it might accumulate another 5-10 minutes of delays by the time it gets to Natick Center since it is operating at reduced speed due to the "less than clear" signals it is seeing.
All of this means that it is really too early to judge the new 5/22/2017 schedule, since the construction related speed restrictions are skewing the on-time performance.
Keolis intended to send "stone trains" last week to dump new ballast along the last stretches of the tie replacement project in Southboro. Once that ballast is tamped into place, the speed restrictions should be lifted after 24 hours.
There are or were a number of other issues that have resulted in speed restrictions and prevented us from seeing the true performance of the new schedule.
2A) Intervale Road bridge: In March 2017, during the routine quarterly inspection of all commuter rail lines, the geometry train discovered an issue at the Intervale Road overpass in Weston (between Route 128 and Wellesley Farms station). This resulted in a 30 mph speed restriction for every train passing over this bridge. Some work must have been completed at the bridge, but the speed restriction was not removed until the geometry train reinspected the line in early June 2017. The Intervale Road bridge will be replaced in the next year or so.
2B) CP 6 construction: "CP 6" is the new interlocking (set of switches) being constructed in Brighton alongside the Mass Pike near the original Staples store location (between Brooks Street and Parsons Street). This interlocking is being built to replace CP 4 which was removed & displaced by the Boston Landing station construction. At various times over the past few weeks, there has been a temporary stop sign for trains on track 2 (the AM inbound track) near this construction area. The stop sign has not existed every day, but when it is in place, trains typically stop there momentarily or proceed slow after receiving permission to go past the stop sign by the work foreman. The CP 6 construction is not scheduled to be completed until well into the Fall, so intermittent speed restrictions may plague us for a few months here.
2C) Heat kink near Worcester station: Early in June, a heat kink was found on the track leading into Worcester station (what's a heat kink? see the picture and explanation here). The kink was repaired without disrupting service to Worcester station, but the entire section of track needed to have ties replaced and railbed resurfacing done. This resulted in a 10 mph speed restriction for 1-2 miles of track immediately east of Worcester. Obviously this is devastating to all trains leaving Worcester, since the speed limit - even on the controlled siding leading to the station - is usually much faster than 10 mph. Keolis intended to complete the work in this area by Friday 6/23/2017.
2D) Speed Restriction near Boston Landing: In early June during the same inspection which cleared the Intervale Road speed restriction, the geometry train discovered a problem near Boston Landing, resulting in a temporary speed restriction there of 30 mph. Keolis has indicated that work will be executed to repair the problem and the speed restriction there should be removed by Friday 6/30/2017.
Hopefully once the majority of these speed restrictions are lifted in the very near future, the on-time performance of our line will improve dramatically. I'm ever the optimist...
2E) Dwell Times: However, it must be noted that some astute riders have noticed troubling delays that appear to unrelated to the speed restrictions described above. These observations are related to longer dwell times at certain stations, increasing any delay that a train might have already had. There have been various reasons observed or noted:
i) Reduced staffing on trains, resulting in fewer doors being opened. Fewer doors means increased dwell time while passengers climb through the limited doors that are open.
ii) Delayed opening of doors. Conductors not opening doors immediately when a train stops can result in delays while passengers wait for doors and traps to be opened.
iii) Not all doors being opened. This could be due to conductor behavior OR mechanical problems, but when only one door/trap instead of both doors/traps at a particular boarding point are opened, all the passengers are funneled through half the usual access points, potentially resulting in increased dwell time.
We'll have to keep an eye on these potential problems.
3) Results of the Public Comment period for the new schedule
As described above, a draft schedule was published in December 2016 and public comments were gathered through a combination of meetings, an on-line survey, and via phone and e-mail. In March 2017, the Worcester Line Working Group met to review those comments and finalize the schedule, which is how we arrived at the final schedule implemented on 5/22/2017.
Overall, 282 people completed the online survey, while 6 people submitted separate comments via e-mail or letters. A number of people commented at each meeting (I attended each one). The MBTA aggregated the responses and presented some statistics regarding the responses, but since it was a self-selected survey and not a scientifically valid poll of a randomly selected subset of the ridership, the statistics from the survey don't represent valid statistics for the ridership as a whole. At the Worcester Line Working Group meeting in March we all recognized that fact and used the statistics for what they were - a method to generalize the responses from the survey.
We discussed a number of topics during the March meeting and I'll explore the most relevant topics here:
3A) AM commute: Most of the survey responses indicated "more or similarly convenient" for the AM commute on the draft May 2017 schedule. There were some complaints about the minor changes in departure times for some trains. There were no recurring complaints or issues which motivated the Worcester Line Working Group to reconsider the draft AM schedule, so there were no changes to the station stop patterns on the draft AM schedule. Some minor tweaks were made to certain departure and arrival times after the full modeling process was completed, but no station stops were added or removed from the draft schedule published in December 2016.
3B) PM commute: The survey responses for the PM commute on the draft May 2017 schedule were split between "similar, less, and more convenient." There was a wide range of comments and reasons for this variation.
The draft May 2017 schedule combined what had been separate express and local trains (departing South Station at 3:30 PM and 3:40 PM respectively) into one full length (Boston to Worcester) local train. As previously explained (see item #8 here), there were a number of reasons for making this change. Approximately 40 comments / survey responses indicated dissatisfaction with this change. We discussed these comments at length at the March meeting and decided to stick with the plan to combine the two trains.
A number of responses commented on the lack of outbound service between 2 PM and 3:30 PM.
Additional responses indicated concern regarding the change in departure times for PM trains, primarily focused on the change of some trains to a 5 minute earlier departure time. Passengers expressed concern that they would not be able to leave work and get to the trains at the new departure times. However, these comments were not of a sufficient volume to prompt the Working Group to reconsider the overall PM schedule. We also discussed and recognized that people who would be pleased or neutral with the proposed time shift (which could allow people to get home earlier in time for day care pickup or other evening activities) might be less likely to submit comments (on the general theory that people are more apt to complain about negative issues).
3C) Boston Landing: The comments received and decisions made regarding the service schedule for Boston Landing are discussed in this blog post (see item #1).
3D) Heart To Hub: There were a wide variety of comments received on the Heart To Hub, and it occupied a considerable amount of the discussion at the March 2017 meeting of the Worcester Line Working Group. The most publicized comment was a letter from Worcester Mayor Joseph Petty. I responded to that letter in this blog post. The letter was discussed but it was recognized that there is no way to accommodate a "better" time for the Heart To Hub with the current infrastructure.
Twenty three comments about the Heart To Hub were received from riders who indicated they used other trains. These additional comments included requests to change the times (similar to Mayor Petty's letter), to NOT change the times, to add more stops, to NOT add more stops, and to eliminate it completely. None of the themes from these comments generated enough consistent ideas to lead to any particular idea being considered by the Working Group.
The draft schedule published in December 2016 proposed adding Ashland & Framingham to the PM outbound Heart To Hub (see item #9 here). Three comments (including the letter from Mayor Petty as a 'comment') were received in opposition to this idea and three comments were received in support of this idea. At the March 2017 Working Group meeting, this topic was discussed at length. We revisited the reason for adding Framingham and Ashland, which was an attempt to increase ridership on this train which has had very low ridership. Passenger counts indicate approximately 75 passengers use the PM Heart To Hub, and this number is not increasing - it has remained consistent since September 2016.
The Worcester Line Working Group agreed to remove Framingham and Ashland from the PM Heart To Hub with the expectation that there may be other ways to increase the ridership on this train, including targeted marketing efforts.
3E) Express trains: 38 people commented on the express trains. It is very easy to summarize those comments: everyone wants an express train that goes to/from their station without stopping anywhere east of their station. That comment was heard from passengers at every station from Ashland to Wellesley. I had a particularly interesting back-and-forth on this topic with a fellow commuter at the Natick public meeting. She was adamant that all express trains should stop at Natick Center, and she did not seem receptive to the idea that adding stops to express trains would a) increase overcrowding on them, or b) increase the trip duration for passengers west of Natick Center. My thoughts on where express and local service should start and end are summarized at item #10 here, and none of the comments I heard or read (during this public comment process or at any other time) have led me to reconsider my opinion that West Natick is the easternmost station that should get express train service. [The comments section below remains open for all you Natick Center passengers to make your best argument for express service!]
It is my hope and expectation that we'll keep the Worcester Line Working Group intact and use it as a forum to discuss future schedule and infrastructure changes.