Today’s Headlines

  • Riders Abandoning Unreliable Subways and Buses in Droves (News, AMNY)
  • Meanwhile — Hey You, Check Out This Shiny Streetcar (News, AMNY, Bklyner)
  • Greg David: Congestion Pricing “the Right Thing to Do for Poorer New Yorkers” (Crain’s)
  • Someone Tell NY1 There’s Always an Adjustment Period After a New Street Design Goes Live
  • Hit-and-Run SUV Driver Critically Injures Man in Crosswalk on 14th Street at Union Square (News, Post)
  • Man, 69, Dies of Injuries Inflicted By Driver Two Weeks Ago; Police and Press Blame Victim (News)
  • Amtrak Will Take 3 Penn Station Tracks Offline for Repairs in January (AMNY)
  • Williamsburg Residents Ask Why MTA Went With Polluting Diesel Buses for L Replacement Service (NY1)
  • Bricks Always Loose on Windsor Terrace’s Main Street Sidewalk (Bklyn Paper)
  • LeBron and the Cavs Paid Subway Tribute (AMNY) and the New York Transit Gods Were Pleased (AP)

More headlines at Streetsblog USA

  • sbauman

    Scheduling to theoretical limits is foolhardy.

    The purpose of calculating the theoretical maximum service level is to use it as a basis to measure how well existing facilities are being utilized.

    The schedule needs to provide an ample buffer to capture random events that delay service a bit.

    That is usually handled by scheduling additional dwell time at each station.

    The signal system has, in fact, been substantially modified since the 1990’s to rectify this problem -[random events that delay service]

    A signal system’s purpose is to avoid collisions and to keep trains within the speed limit. That’s called train control. CBTC stands for Communication Based Train Control. What you described (keeping trains on schedule) is something called train supervision. The form of train supervision used by subway system is too crude to be of any use. This includes the old manual systems used by most lines and Automatic Train Supervision systems in use on the IRT 1-6 lines and the L.

    an increase in stopping distance will reduce capacity,

    The amount of capacity reduction depends on whether on is looking at emergency or service braking. In either case the amount of maximum service level degradation is small because its the braking distance divided by the travel velocity. This amounts to a couple of seconds not tens of seconds.

    Short-turns can be a useful strategy, of course, but they’re best avoided if a significant share of the ridership is to or from points past the short-turn point, since in that case they can lead to significantly variable loads between trains.

    Let’s assume that the planners and schedule makers know what they are doing. (That’s a whale of an assumption.) The penalty for not using short turns is requiring more equipment to provide the same service level where it’s required.

    So, exactly as I said: “How often that schedule was achieved I don’t know, and I doubt you know, either.”

    Here’s a link to BOT’s reported OTP from that era;orient=0;size=100;seq=47;attachment=0

    There’s not much wiggle room with for the IND with an OTP of 99.22% in 1949. The trains operated on time.

    But, as you may be aware, there was an incident on the Williamsburg Bridge in 1995 that suggests that the signaling in that area may not be identical today to how it looked in 1949.

    First, in 1949 the OTP for the BMT wasn’t quite up to the IND (99.19% vs. 99.22%). That still did not leave enough wiggle room to abandon anything like the 5% of rush hour trips, as the MTA commonly does today.

    Second, differences in grade timers must be at station entrances to have any effect on maximum service levels. Any changes to the signals on the WB, especially in the vicinity of the crash, would directly change only travel time – not train frequency.

    The increased travel time would require more trains to operate the same frequency. That’s not a signal problem.

    That [grade crossings should not limit capacity] only works if (a) frequencies are identical in both directions (or one is a multiple of the other) and (b) the trains in opposite directions arrive precisely as scheduled. Perhaps (a) is the case. I can ensure that (b) is not the case. All it takes is one person to hold the doors for a few seconds at any of the previous 27 stops to throw off the scheduled meet.

    (a) is not the case. Numerous switching conflicts are scheduled. Having a train supervision system and a schedule that includes dwell times should insure that a couple of seconds delay at one stop will be made up by the next. Also, door holding will be far less frequent when trains come every 2 minutes or less rather than every 4 or 5.

    Demand on the M is heavier coming from Williamsburg than from Queens Boulevard in the morning

    That data is 2 years old. MTA Board meeting was held in Oct 2015 meaning survey was held well before then. Since that time, several large buildings have been opened in LIC and several more are on the way. One of the MTA’s big problems is its inability to anticipate and/or react to long term demand shifts.

    Also, think about what this does to your state goal of “balanced merges everywhere.”

    I have. The need for balanced merges in Brooklyn is for the M vs. J/Z is at Essex St (southbound) or Marcy Ave (northbound local-express) or Myrtle Ave (norththbound local-local). If J/Z operates at 12 tph and M operates at 15 tph, 3 M’s need to short turn in Manhattan.