Jump to content

Stephen Bauman

Veteran Member
  • Posts

    18
  • Joined

  • Last visited

Everything posted by Stephen Bauman

  1. 7 from 34th to Main St 4/10/19 "00:39:06.419214";139.139961017714 5/15/19 "00:39:09.141631";130.222427854399 7X from 34th to Main St 4/10/19 "00:32:38.225806";66.3454494850382 5/15/19 "00:32:50.769231";81.2847558799406
  2. Here are the run times from the supplemented from 4/10/19 7 4/10/19 "00:38:33.84";135.902658289382 7X 4/10/19 "00:33:37.142857";74.907608597418
  3. Here's the Manhattan bound data for 5/15/19 from the supplemented schedule that I downloaded yesterday morning (5/10/19). 7 5/15/19 "00:39:04.235294";161.586936722929 7X 5/15/19 "00:33:40.465116";79.0705980022824 As you can see, there's usually a fairly close relation between the supplemented and GTFS schedules. There is an archive for supplemented schedules that dates back to September 2017. My experience has been that substantive differences between the supplemented and GTFS have been few and far between.
  4. A new GTFS timetable was released on May 9th. Here's a comparison of Main St to 34th St scheduled running times. 7 to Manhattan 5/18/16 "00:38:30.75";139.176673489183 5/17/17 "00:38:30.860656";137.501504103933 5/16/18 "00:38:33.84";135.902658289382 5/15/19 "00:38:34.5";138.059287247786 7X to Manhattan 5/18/16 "00:33:37.142857";74.907608597418 5/17/17 "00:33:37.142857";74.907608597418 5/16/18 "00:33:37.142857";74.907608597418 5/15/19 "00:33:40.465116";79.0705980022824 It's date, average running time in "hh:mm:ss" and standard deviation in seconds. The GTFS schedules don't reflect any run time reduction as a result of CBTC's introduction (with or without ATO).
  5. There is a theoretical maximum TPH, that depends on the emergency braking rate. How close does CBTC approach this limit? How close does the existing block system approach this limit? There's a difference between an algorithm and its hardware implementation. Digital logic's modern hardware implementation does not use relays nor the railroad industry's perverse vital circuits concept for reliability. It uses programmable logic controllers (PLC). The PLC's are commodities whose cost varies directly with the number of I/O ports rather than logic complexity. Thus, adding ubiquitous, accurate logical timers is a nominal extra expense. There are approximately 16,000 signals in the system. This comes to a $2.5M replacement cost per signal with the projected $40B Fast Forward budget. The railroad industry's rule of thumb is about $1M per signal, using vital relay technology. It would probably be half that, if PLC's were used.
  6. What they are not saying are: the service (TPH) levels that CBTC will make possible; the service levels they plan to operate with CBTC; the service levels they currently operate; the service levels that are possible with the existing signal system; and the service levels predecessor agencies actually operated with this signal system.
  7. This is a quote from the TTC report that was referenced by Mr. bulk88, above. It's a misconception. Let's calculate the headway between trains, with the follower traveling at the minimum safe distance. Let's assume the leader and follower are traveling at the same velocity V. If the follower's emergency braking rate is a, then the follower's emergency braking time is (V/a). This is the minimum time between the rear of the leader and the front of the follower. Let an observer at a point witness the time interval between the passing the front of the leader and the front of the follower. That time is the headway, T. It consists of the time for the leader of to pass the point plus the emergency braking time. If the leader is traveling at velocity V and has length L then T = (L/V) + (V/a) where (L/V) is the time for the leader to pass the point and (V/a) is the minimum safe braking time for the follower. This expression for the headway has a minimum when V = SQRT(aL). The headway at this velocity becomes T = 2 * SQRT(L/a). Plugging in the nominal values of 3 mph/sec for a and 600 ft for L, yields an optimum value of V = 51.4 mph and a minimum value T = 23.3 seconds. Thus, the strategy of operating speeds slower than 51.4 mph with shorter distances between trains is not an effective strategy for decreasing the headway between trains.
  8. The operational problems on the Canarsie Line might be "solved". However, it's been shoved onto the public. Weekday ridership is much greater. The proposed surface transportation enhancements would not have been sufficient. It would have been worse than the proposed 3 tph for weekends with no surface transportation enhancements.
  9. You're right regarding the 15 minute gap at the Lorimer switch. My mistake. I did try to estimate the accumulation of incoming passengers during the 24 minute gap at each of the stations. I used the turnstile data that's available on the MTA's website. It's reported in 4 hour segments. I just divided the entry count by 8 to guess what a 30 minute accumulation might be. The counts at Bedford are between 543 and 691 from 11am to 7pm on Saturday and Sunday. I'd prefer to have the extra platform space to avoid any mishaps. None of the single line stations between Bedford and Canarsie approach these figures. Graham is the next closest with an accumulation around 150 during these hours. I wonder whether a shuttle is worth the effort, given the sharp drop. The next busiest single line station is 1st Ave, with accumulations around 300 from 11am to 11pm. It's a mistake to make the second busiest station on the line exit only.
  10. I'm nixing the possibility of running a shuttle between Bedford and points east. The switch will see trains crossing it every 90 seconds in one direction or the other. That leaves not time to sneak in a shuttle. This is not a 24/7 operation. It's temporary for midnight and weekend hours. At other times its normal operation. Therefore any temporary overbuild at 1st and 3rd would have to be removed and set up within a few minutes. If you have a low cost implementation in mind - great. I haven't.
  11. Actually there's another problem. The scheduled times during single track operation were 10 and 8 minutes. The extra 3 minutes over the 7.5/8.0 means that only 10 tph operation is possible. The dead time for each direction would be raised to 24 minutes. I'd have to check the real time tunnel crossings to see if there were 3 minutes padding over the best times to get to 12 tph. The stations have to be monitored to prevent overcrowding. I'd lock the turnstiles, to prevent more people from entering an already crowded platform. Maybe Paris could lend us some of the portillions automatique, if they still exist. There are electronic people counters, so personnel can be used for more important work. The major effort is to prevent dangerous situations on the platforms. I have a couple of ideas to add safety and capacity to the platforms. I would use an empty train on the unused track at Bedford to extend the platform width by 10 feet. It's vital to separate exiting and entering passengers from each other. It might be possible to rope off exiting and entering corridors on the platform. The entering passengers would be waiting on the empty train across the platform for exiting passengers to leave the train. The side platforms at 1 Ave and 3 Ave are easier. CBTC under ATO is supposed to be able to stop within 1/8 inch of its mark. I'd put it to the test but I can relax the tolerance a bit. I'd park a train on the unused track. The train would be equipped with a gangplank on the side opposite the station. The gangplank would extend to the running track and be used to load passengers. (On second thought, the gangplanks would remain in the station between the rails. Their length would be such that it would extend to each track as an extra platform. They would not interfere with trains entering or leaving.)Trains would enter these stations and open their doors on the station side for exiting passengers. The train would then open the opposite side for incoming passengers. The doors opposite the platform on the "waiting room" train would be used as platform doors. Both 1 Ave and 3 Ave are straight platforms, so there should not be clearance problems. There would be no need for the moving platforms used at Union Sq (IRT) or South Ferry Loop. I think the strategic placement of these empty trains should solve safety concerns, as well as permit 30 second dwell times.
  12. Both 1 Av and 3 Av are narrower side platform stations, whereas Bedford is a wider island platform. With 20 instead of 4 minute headways, the MTA rightly feels that overcrowding on the narrower platforms will be dangerous. It will be just as dangerous at Bedford. 20 minute headways during weekend days should be a non-starter. There are options, that NYCT isn't considering.
  13. The only info I have is what's publicly available from the GTFS-RT feed. I'd assume looking into the causes for the late starts should be one topic for the management investigation I suggested. All the information regarding scheduled and actual arrivals and departures for all the station is contained in the linked spreadsheet. I'd suggest anyone study it to gain insight into NYCT operations.
  14. I had worked this out with travel times of 7 minutes from Union Sq to Lorimer and 7.5 minutes the other way. These were the minimum scheduled GTFS running times, when I figured this strategy out back in 2016. I can re-work it with the scheduled travel times from the Jan 15-18 midnight hours. It should not make much of a difference. The nominal intermediate station service level capacity is 40 tph, including 30 sec dwell time per station. That's independent of signal system and has been achieved in practice by both Moscow and the BOT. The service level capacity is usually limited by the terminal stations. This is true with the 14th St Line. If trains are operated 6 at a time in each direction @ 40 tph the time interval between the first and last train will be 5 x 1.5 min or 7.5 min. The total travel time for 6 trains through the tunnel Brooklyn bound would be 14.5 minutes and 15 minutes Manhattan bound. This adds up to 29.5 minutes for 6 trains in each direction 12 tph. This still leaves the terminals to be considered. Neither 8th Ave nor Rockaway Pkwy can handle 90 sec headways nor direction reversals without delaying followers. The solution is that not all trains would terminate at their respective terminals. Let's assume the Manhattan bound tunnel is operational. On the Manhattan side the first train would terminate at 8th Ave; the second at 6th Ave and the third at Union Sq. The next 3 would do the same but wrong rail to 8th Ave, 6th Ave and Union Sq. respectively. The Brooklyn bound train would be the third train which had terminated at Union Sq. It would be followed by the second, first, sixth, fifth and fourth arriving trains. Each had adequate time to recharge brakes to turn around. Rockaway Parkway is a bit more complicated because the closest crossover to the Manhattan bound track (Q1 to Q2) is between Bushwick and Bway Jct. There are trailing crossovers north (Q2 to Q1) north of Livonia and Sutter, in addition to the diamond crossovers north of Rockaway Pkwy and Bway Jct. The no construction cost option would be to have the first two trains terminate at Rockaway Pkwy, and the next terminate at E 105, New Lots, Livonia and Sutter on the Brooklyn bound track. The return trip would have the first two Rockaway Pkwy trains depart onto the Manhattan bound track, followed by the train that terminated at Sutter using the trailing point switch north of Sutter, followed by the trains that terminated at Livonia, New Lots and E 105 using the trailing point switch north of Livonia. CBTC makes such an out of the box operation feasible because it should be able to wrong rail at maximum service levels. Substantial modification would be needed with a block system which wasn't built with this capability. This would mean 40 tph operation through the Canarsie Tunnels which the MTA claims it can handle only 20 tph without additional substations. However, the 40 tph operation is in only 1 direction whereas the existing 20 tph operation is in both directions. This would require the same amount of peak amperage that the existing substations currently supply. As noted above the cycle time is 29.5 minutes. However, it takes 7.5 minutes for the 6 trains to travel past a single station. Therefore, the worst case wait time scenario would be 22 minutes.
  15. Subchat appears to be down. Here's the link to the spreadsheet that shows the performance of single track operations on Jan 15 - 18 of last week. It covers the time period from midnight to 5am. The 3 columns for each trip are RT0 which is the initial schedule posted on the real time feed. It's usually posted approximately 30 minutes before departure. The post time is given above the RT0 identifier in the header. If the post time is much closer to the departure, it means the dispatcher re-arranged the schedule. RTN - The schedule is displayed and possibly modified every 30 seconds until the trip is completed. Only the remaining stops are given in each update. The RTN data represents the last update for each station. This is the data that's used by the countdown clocks at each station. POS - There's also vehicle position data that notes the status to the next station. The arrival time is given by the timestamp part of the position message for the first STOPPED_AT status. It generally corresponds to the time when the station countdown clock displays a steady orange 0 minute message. It's approximately 20 seconds before the train actually stops within the station. The departure time is sending time for the last STOPPED_AT status position message. It's generally when the doors close. I've repeated the schedule showing only the RT0 columns below on each sheet, to make it easier to follow what was planned. The 15th and 16th show tunnel travel times of 7 and 10 minutes with Union Sq to Lorimer being the longer. That's a total of 17 minutes, leaving 3 minutes buffer time for a 20 minute headway. The 17th and 18th show tunnel travel times of 10 and 7 1/2 minutes with the Union Sq to Lorimer being the longer. This leaves a 2 1/2 minute buffer for 20 minute headways. If the buffer times are assessed equally between Brooklyn and Manhattan bound trips, each trip needs to be within 75 or 90 seconds of schedule to avoid single track delays. An inspection of the POS column for the Canarsie departures reveals this criterion was often violated. Further inspection of the late start trips from Canarsie to Lorimer reveals that the CBTC/ATO/ATS combo was not particularly good at gaining lost time during the trip to Lorimer. Good management would investigate the reason behind the late starts and missed connections at the Canarsie Tunnel. The investigation's aim should be to avoid repeats. 20 minute headways during the midnight hours would not impose a hardship because it roughly duplicates the current service levels. It's a different story for weekends. 15 tph operation is scheduled for weekends. NYCT is not known for operating too many trains, when demand does not exist. I'd assume these trains are packed, although I've not sampled them. It should be possible to operate 12 tph, with single track operation through the Canarsie Tunnel under the existing CBTC and power constraints. It would take a lot more operational discipline than NYCT exhibited during the 3 tph operations this past week.
  16. You have to define quantitative performance criteria that CBTC must meet, in order to answer that question. First, does CBTC meet all the criteria? Next, are there other technologies that can meet the same criteria? If the first is true, then CBTC would be "sufficient" in a mathematical sense. If the second is false, then CBTC would be "necessary" in the same mathematical sense.
  17. It's fairly easy to calculate the effect of train length on max TPH. Headway is the inverse of TPH, with a change from hours to seconds. 40 TPH translates to a 90 second headway. Part of the headway calculation is the time it takes a leader to clear a block for the follower. The leader will usually achieve cruising speed before the block between him and the follower is cleared. If we assume that cruising speed is 30 mph (45 ft/sec) and train lengths were increased from 400 to 600 feet, then it would take an extra (600-400)/45 or 4.4 seconds. Thus the headway would be increased from 90 to 94.4 seconds and the maximum TPH would be decreased to 38 tph.
×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use.