Here are my observations from 13 Apr 2002
I was interested to see what the TA's simulated 30tph operation
would be. I witnessed this mornings simulation of a morning rush
hour. I also witnessed a couple this past week to establish a
baseline. I figured that any organization that currently does not
know how frequent their past service was might not know what its
present service is. I know the figure is not 24 tph based on TA
press releases proclaiming 29 and 30 tph operation over the past
couple of years.
There are about 5 key locations that should be monitored
simultaneously. I chose to observe Manhattan bound traffic at
111th St. This permitted me to observe the operation of both
express and local operation in sufficient detail. Service is
divided into express and local service with a merge at 33rd St.
This is a 1:1 merge (equal service both the local and express).
Local service is further divded at 111th St with half the locals
originating and terminating at this station. The remainder of
locals and all the expresses originate at Main St. I am assuming
that no morning trains originate or terminate at Willets Pt.
The steady state rush hour starts with the first express at 6:40
and ends around 8:47, when the last 111th St local enters
service. This is a period of roughly 2 hours. The round trip time
is slightly over 1 hour, so most trains make 2 round trips.
I recorded the arrival times for all trains from Main St. I
recorded both the arrival time and the time of the dispatcher's
signal for the 111th St locals.
The current TA timetable lists the difference in running time
between the local and express as 6 minutes. An older timetable
showed a 5 1/2 minute difference. This means that both the local
and express should leave 111th St at the same time to avoid
merging conflicts at 33rd St for 30 tph operation. They did not
do this which makes me believe that there was a fair amount of
merging problems at 33rd St.
It would appear that many of operating personnel were not
familiar with this line. Four conductors of the 15 111th St
locals did not wait for the dispatcher signal before closing the
doors. Another of the 111th St locals did not open its doors for
3 1/2 minutes after its arrival and did not open until 3 minutes
after the dispatcher's bell sounded. This snafu did elicit a lot
of motion on the part of the many TA personnel that were on the
platform. I did not notice any 111th St local leaving early
during the previous days' obervations.
One major problem with the 30 tph operation is that they ran out
of trains around 8:30. This increased the gap to the extent that
two locals from Main St skipped 111th St. This resulted in
intervals between trains of 14 and 16 minutes. There was a single
non-stop incident on Friday. This resulted in an interval of 11
It is extremely difficult to do any number crunching without an
accurate published schedule. It is also extremely unlikely that
it is followed, if it were available. The dispatcher's clock at
Main St does not show seconds. It was off by 10 seconds from NIST
time. (I do set my watch to an NIST standard before taking
I have done some elementary number crunching. I've taken a moving
window of the number of Manhattan bound trains that passed 111th
St from 6:50 to 9 AM. I've moved the window in 1 minute steps.
I've chosen 3 widths for this window: 1 hour; 30 minutes and 15
minutes. I've taken the average for these windows to come up with
a single service level count. I've also taken the max and min.
I've used the arrival times for the trains that originated at
Main St and the dispatcher's bell for the 111th St trains. The
results are as follows: for previous 11-car operation for the 1
hour window the (avg[eff 1 hr rate], max, min) are: (25.61, 28,
23). For the 30 minute window: (12.68[25.36], 14, 11) and for the
15 minute window: (6.33[25.32], 8, 4). For 10-car operation and
"increased service levels" the same results are: (28.06, 31, 24),
(13.65[27.3], 16, 9), (6.66[26.64], 10, 4). N.B. these figures
calculated assuming that the trains that skipped stops did stop.
What is striking is the increased variability with the attempt at
30 tph operation. One should expect a deviation of only +/- 2
(express and locals leaving at same time). The fact that they ran
40 tph for 1/4 hour when one should be operating only 30 tph
means that people were leaving too soon. This may have
contributed to the train shortage 1 hour later. N.B. decreased
variability is required for increased service levels.
Was an equivalent amount of service provided with 10-car trains?
Passengers are effected by the system's performance during the
last quarter hour or less. The effective number of cars/hr for
11-car operation was 6.33 x 4 x 11 = 278.52. The effective number
of cars/hr for 10-car operation was: 6.66 x 4 x 10 = 266.4. for a
declne of 4.4%.
I would definitely catagorize the delay of nearly 6 minutes in getting operating personnel to operate a departing train to be part of the TA's lack of operational ability. There were about 5 supervisors on the Flushing bound platform There weren't any on the platform where the trains were supposed to leave for Manhattan.
2. Has the time for a trip, say, the express run from Times Square to Main Street remained constant? If not, then what has the change been?
When the R33-36's were introduced both express and local running times were reduced about 3 minutes. The local became 30 minutes and the express was down to 25 minutes.
3. Has the minimum distance between trains at all points remained (significantly) the same since the decrease in service? If not, then what has the change been? In terms of signalling, I guess the question would be, has the number of blocks separating trains remained the same, i.e., are those safety standards constant? This is prompted by the reported changes the MTA made following the 1990s that they were bringing to "modern standards" the distance between trains.
There has been no significant change in the signal system, since it was installed in the early 1950's until the 2002 test.
4. On that last point of "modern standards" of safety I admit to some confusion. The reasoning as I understand it (which may be wrong) is that trains are heavier and take longer to stop than they once did.
First off, we are talking about the same 40 year old equipment. You will have to ask the TA, why both old and new equipment takes longer to stop (time and distance) than the prewar fleet, including the D-types, which were the system's heaviest.
On the former, though, trains seem to be lighter than they were (the NTSB report for instance that the MTA purports to rely on in making a long term change to CBTC mentions that in the late-1960s a prime design requirement of trains was that they be lightweight).
The reason for wanting lightweight equipment is to reduce the amount of power required to propel them. This translates into lower operating costs because of reduced electrical demands.
I cannot wite on the latter, although I would be (somewhat) surprised to learn that braking had not improved any in over fifty years (IIRC, they used the R-1 or R-10 or some other like model to establish the mean distance between signals, which is they say, too low for today's faster, slower to stop and heavier trains to consistently meet). Can anyone explain what seems like a paradox?
Actually the R10 was the first trainset that reduced emergency braking capability.
FWIW, the best emergency braking capabilities were on the prewar Bluebirds, which used track brakes as one of its 3 braking systems. The BIE was 7 mph/sec or more than double today's klunkers.