Home · Maps · About
Home > SubChat
 

[ Read Responses | Post a New Response | Return to the Index ]
[ First in Thread | Next in Thread ]

 

view flat

Re: Canarsie CBTC

Posted by Jeff H. on Fri Jun 17 03:52:02 2005, in response to Re: Canarsie CBTC, posted by RonInBayside on Thu Jun 16 09:10:27 2005.

edf40wrjww2msgDetail:detailStr
After you fill in your own understanding of CBTC (hhey, and ours too, OK?) then you can intelligently discuss the
alternatives. Not before. And you have not done that yet.

Yeah I have numerous times but you weren't paying attention.

In a nutshell, this is my understanding of how CBTC works,
from memory, without quoting from things on the internet:

Railway signaling (as it pertains to the NYC transit system, so
we're not talking about things like hump yards and rock slide
detection) is concerned with two primary things: safe spacing
between trains and "interlocking" (the coordination between the
safe movement of trains and the movement of switches). CBTC
does not address interlocking, so let's move on.

There are numerous approaches to maintaining safe distance between
trains. We'll consider only those systems that provide Positive
Train Control, e.g. the trip stop mechanism currently used, as
opposed to those which rely solely on the operator's observance.
In order to keep trains separated, it is necessary to determine
their position. All signal systems to date have used track circuits
to do this. I will not explain what a track circuit is as I
think it is common knowledge.

In addition to knowing a train's position, it is necessary to be
able to make some statement about its speed, and its maximum safe
braking distance at that speed.

In a wayside ABS such as used in NYCT, determination of position can
only be made at the granularity of the track circuit boundaries
(the insulated joints). Likewise, speed through a block must be
assumed to be the maximum fleet speed (this approach works because
NYCT does not mix equipment with radically different top-end speeds
in service as a mainline RR might), OR if a "timer" is used, then
speed has been demonstrated to be some lesser number. Commands
to the train crew as well as enforcement commands (stop arm up)
likewise can only be transmitted at track circuit boundaries.

In NYCT's ABS system, the variables are very simple and in fact
reduce to booleans. All of the logic is performed in the wayside
signal locations. There is no car-borne equipment except for the
trip cock. Interoperability concerns are limited to the position
of the trip cock on the car.

Another system which has been successfully used in thousands of
miles of track, both mainline and transit, is Automatic Speed Control
with coded track circuits. In this system, the wayside logic can
establish one of a fixed set of allowable speeds through a block.
This command is transmitted through a very simple and robust
manner: by modulating the track circuit energy at a very low
frequency (generally 1-7 Hz). The car-borne equipment is relatively
simple and consists of an inductive loop which picks up the
modulation, some passive filter components, some relays, and a
speedometer. Since the train is responsible only for measuring
speed, not position, the speedometer accuracy is not critical.

With ASC, although the commands are transmitted at discrete points
along the railroad (again, the track circuits), they are being
transmitted continuously, as opposed to the intermittent commands of
a wayside color light system with trips.

ASC provides positive enforcement of speed and therefore the wayside
system "knows" the speed of a train through each block, and can
allow following trains to "creep up" without having to add a bunch
of shorter blocks and "station timers".

ASC achieves good cross-vendor interoperability because the "protocol"
is so simple one can analyze it with a Simpson 261 multimeter.


Many people have a mental equation in their heads that

OLD SCHOOL SIGNALS == RELAYS

That isn't the case. Systems have been installed for several
decades now which use microprocessors to replace 99% of the relays
in both wayside ABS, interlocking, and coded track circuit ASC
systems. So next time somebody gives "eliminating relays" as
an advtange of CBTC, keep that in mind.

The paradigm shift of CBTC is to make the train responsible for
knowing not only its speed, but also its position. There are
a number of technological ways to do that, but they all boil down
to measuring elapsed distance (and with a little d/dt, speed)
and periodically checking that against fixed beacons.

Another key element of CBTC is that the train estimates its own
safe braking distance. This requires tight integration with the
propulsion and braking packages (and therefore makes retrofit
of "legacy" equipment difficult).

Under CBTC, the train is in constant communication with a wayside
control element. This communication is digital and follows a
complex protocol with many thousands of lines of code to drive it.
Again, there are many possible technologies which can be used in
the communications link and, as of 2005, no consensus as to what
will become the standard, in terms of the physical interface,
the data link interface, or the network interface.

The wayside zone controller maintains communication with adjacent
zone controllers and is able to hand-off as the train passes from
zone to zone. The controllers constantly monitor the position,
speed and braking profile of all trains and constantly transmit
commands back to the train giving their allowable speed. The trains
enforce the speed either with speed control (motor cutoff and
penalty applications) or completely automatic operation.

The elegance of CBTC is that it removes the intermediate layer
of track circuit and gets right down to the essentials: speed,
position, braking. In theory, this approach could allow trains
to operate as quickly and as closely together as is safely possible,
without any loss of efficiency from the signal system.

Another advantage of CBTC is the reduction in maintenance of
wayside elements. Although CBTC has plenty of wayside equipment,
none of it is in the roadbed with trains passing over it (well,
that's a stretch, but basically correct).

That's the theory of CBTC. [CBTC does not provide interlocking,
and I have not discussed how the CBTC system interfaces with
interlocking systems]

People have this idea that Canarsie trains will be flying along
at top speed only a few hundred feet apart, and that tph capacity
will be greatly increased. Unfortunately, that is not the reality
of the Canarsie CBTC implementation, and I mean yes, even when and
if it is working 100% according to spec.

Basically, on Canarsie, the track is broken up conceptually into
a number of "virtual blocks" which are fixed in location, and are
comparable to the existing blocks in size. You essentially are
getting exactly what ASC with coded track circuits gives, minus
the insulated joints (and impedance bonds, since single-rail coded
track circuits don't work out so great), plus a lot of very
complicated and proprietary wayside and car-borne equipment.

Now, I say to you, what's the relative value proposition? Will
the maintenance savings on the insulated joints outweigh the
higher install cost of CBTC, the need to hire more expensive
consultants and specialists, and the future headaches of
incompatibility?

Responses

Post a New Response

Your Handle:

Your Password:

E-Mail Address:

Subject:

Message:



Before posting.. think twice!


[ Return to the Message Index ]