On 2008-12-04 18:31:46 +0000, Tom Anderson said:

On Thu, 4 Dec 2008, Robert wrote:

On 2008-12-04 16:24:15 +0000, Tom Anderson said:

On Thu, 4 Dec 2008, Andrew Heenan wrote:

(And please don't tell me there's not one station on the Eastern that
is more appropiate than Shenfield - or I, and many others, will cease
to believe a word you say.)

As it happens, i also think there's no better option than Shenfield.
The GEML is four-track to Shenfield, and has two two-track branches
beyond that. That means you can run Crossrail as a stopping service to
Shenfield with one pair of tracks entirely to itself, and leave the
other pair for non-stop long-distance services, with no possibility of
performance pollution between the two. Running those Crossrail trains
beyond Shenfield supplies residents of those towns with a stopping
service into London which they simply won't use. Making some of the
Crossrails non-stop on the fasts to points beyond Shenfield, and
filling in the deficit on the slow lines with Liverpool
Street-terminating trains, throws away the advantages of segregation.
Turning some of the Crossrails off short of Shenfield - say up the West
Anglia, to suburban destinations or Stansted, means taking trains away
from the stations towards Shenfield, which means a net reduction in
service on an already overcrowded line. So, we have one useless option,
one impractical one, and one actively harmful one.

I look forward to hearing your suggestion.

To balance peak hour loadings between trains running on the same Line
on the RER in Paris, the stopping patterns are varied. For example on
the western arm of the Line A, most trains ran through to the terminus
at St. Germain-en-Laye, but some turned back before the end of the line
at Le Vesinet-Le Pecq. The longer distance trains tended to skip some
of the stations nearer Paris which were covered by the trains which
turned back early. All of the trains stopped at all of the stations in
the central section. This was all done on a 2 track railway and it
seemed to work very well. Outside the central section I would suggest
that not all the Crossrail trains should be all-stations.

The only point to this that i can see is to guarantee that passengers
on the inner section have a chance to get seats; if all trains ran the
whole way, passengers from the outer section would all get a seat, and
get all of the seats, leaving none for the poor inner sectioners. The
flip side of this is that some people coming from the outer section
will have to stand, despite having further to travel than anyone from
the inner section.

That was, I think,the main reason. The maximum journey length was about
25 minutes from Étoile. The lack of seats problem arose mainly in the
evening peaks when trains leaving the central section were full and
standing. I lived there about 12 years ago and I seem to remember that
there was a pattern of 3 trains every 10 or 12 minutes down the St.
Germain branch. Also using the central section tracks on the Line A
were the dual-voltage trains to both Poissy and Cergy which diverged at
Nanterre-Prefecture. So there were trains every 90/120 seconds or so in
the central section which had automatic train control. In the opposite
direction, as St. Germain was a terminus, you could practically always
get a seat in the morning peak. The St. Germain trains were always 3
sets of 3-car emus, I forget the class number but they were built in
the 50s; the Cergy/Poissy trains were modern aluminium bodied 4-car
units running in pairs. The total train length was the same for both
types.


Specifically, the arrangement can't make the long-distance trains much
faster. Since trains can't overtake on a two-track railway (without
passing loops, anyway - do they have those?), then assuming that
stopping trains are all evenly spaced, the most time that a skipper can
save is equal to the interval between stoppers - if it sets out from
the central section just ahead of one stopper, it can reach the
turnback point just behind another one. If the trains all come out of
the core evenly spaced, skippers and stoppers, then the maximum gain is
the time between a skipper and a stopper - half the interval between
stoppers, if they're half and half.

There were no passing loops, except at Le Vesinet-Le Pecq where the
central reversing siding had 2 platform faces, being the insides of the
up and down island platforms. In my previous post I forgot that trains
could also be reversed in 2 bays at Rueil-Malmaison, but that was
mainly to be able to get trains into and out of the maintenance depot
there.

Crossrail is going to run at 12 tph along the GEML, along with another
6 tph of Liverpool Street trains. That's 18 tph, or a train every 3
minutes 20 seconds, that also being the maximum saving a skipper could
make. That doesn't seem like much of a saving over the 36 minutes it
currently takes to run from Stratford to Shenfield, particularly when
compared to the 17 minutes it takes non-stop on the fasts.

3min 20 secs! That's terrible. Modern metros should be able to run 30
tph or more. I've lived in both Paris and Munich - both cities can
manage that frequency. This country is backward!

Bear in mind that this would represent a cut in stopping train service
on that line from 16 tph to 9 tph. If you introduced more Liverpool
Street stoppers to make up the difference, you'd cut the time saving
for the skippers even further.

Another alternative along these lines is a skip-stop service, where
half the trains skip half the stops, and half skip the other half. This
does actually let you get trains from one end of the line to the other
faster, although not a lot faster in practice, i believe. There isn't a
capacity issue, as although the frequency at each station is halved,
the trains which call are each serving half the number of stations. It
does double the average waiting time for a train, but at 18 tph, that's
from 3:20 to 6:40, which is still a reasonable turn-up-and-go
frequency. It doesn't really help you run trains beyond Shenfield,
though.

tom


--
Robert