Dave Arquati wrote:

Don't forget that an interchange (well, an H&C station on Wood Lane) is
already being built at White City, and will have been in operation for a
number of years before Crossrail ever appears in the area.

The other issues would be what to do with Hammersmith Depot and where
Circle Line trains should be stabled.

The stock requirement of the Circle (or whatever succeeded it)
operations would go down somewhat. The remaining stabling arrangements
could be to retain a stub towards Royal Oak along which trains could be
stabled; similarly, perhaps use could be made of the defunct Moorgate
Thameslink branch for expansion of the Farringdon sidings.

The new common S-stock should allow greater flexibility in where to
stable trains, so any more sidings needed could be distributed around
some other depots wherever a small expansion could occur.

--
The S-stock is something, the merrits of which, I have yet to be
convinced. Time will tell if the concept is acceptable.

If one dreams a little further with regard to CrossRail and the
Hammersmith branch: Imagine, if you, will a rebuilt connecting viaduct
at Hammersmith allowing CrossRail to reach Ealing Broadway and
Richmond.

The problem here is the purchase and demolition of all that has been
built in the intervening years. There is also the issue of the many
differing platform lengths en route to the two termini.

However, if there District Line was reduced to an Upminster to
Wimbledon service plus an Edgware Road to Kensington Olympia shuttle,
Circle Line operation would become simplicity itself!

Adrian.

wrote:
Dave Arquati wrote:
Don't forget that an interchange (well, an H&C station on Wood Lane) is
already being built at White City, and will have been in operation for a
number of years before Crossrail ever appears in the area.

The other issues would be what to do with Hammersmith Depot and where
Circle Line trains should be stabled.
The stock requirement of the Circle (or whatever succeeded it)
operations would go down somewhat. The remaining stabling arrangements
could be to retain a stub towards Royal Oak along which trains could be
stabled; similarly, perhaps use could be made of the defunct Moorgate
Thameslink branch for expansion of the Farringdon sidings.

The new common S-stock should allow greater flexibility in where to
stable trains, so any more sidings needed could be distributed around
some other depots wherever a small expansion could occur.

--
The S-stock is something, the merrits of which, I have yet to be
convinced. Time will tell if the concept is acceptable.

I'm not yet convinced by the internal layouts, but the idea of a stock
that can go "anywhere" seems sensible when the future of the SSL network
is hazy.

If one dreams a little further with regard to CrossRail and the
Hammersmith branch: Imagine, if you, will a rebuilt connecting viaduct
at Hammersmith allowing CrossRail to reach Ealing Broadway and
Richmond.

I don't think I'd bother going that far. Taking over as far as
Hammersmith is probably OK because Crossrail serves several of the H&C's
most important destinations. However, it goes nowhere near Victoria and
is a bit of a walk from places like Monument in the City. We know that
Richmonders don't want Crossrail to replace their District service, and
Ealing Broadway will already have Crossrail.

Shared running would also be a very poor idea as the trains would be so
different and there would be the inevitable performance pollution.

The problem here is the purchase and demolition of all that has been
built in the intervening years. There is also the issue of the many
differing platform lengths en route to the two termini.

However, if there District Line was reduced to an Upminster to
Wimbledon service plus an Edgware Road to Kensington Olympia shuttle,
Circle Line operation would become simplicity itself!

Crossrail operation would be a lot more complicated, though!

--
Dave Arquati
Imperial College, SW7
www.alwaystouchout.com - Transport projects in London

On Tue, 1 Aug 2006, Dave Arquati wrote:

Tom Anderson wrote:
On Tue, 1 Aug 2006, Dave Arquati wrote:

Bob wrote:

to turn round of half the trains at Paddington which seems a wasted
opportunity - taking over the Hammersmith end of the H&C removes a busy
junction on the Circle Line with knock on effects for that lines
reliability would seem a low cost no brainer.

I agree, but I think ensuring the central tunnel gets built is most
important - adding on too many extra bits here and there could kill the
whole project. Hammersmith could come later...

Except Hammersmith involves noodling about quite near the Paddington portal
(if you want to have a station at Royal Oak, or if you want to use the
existing H&C fly-under, at any rate), so if you don't at least build in the
possibility at the start, you might not be able to do it later.

I'm not sure it would be worth keeping Royal Oak in this scenario
(gasp). Its catchment area overlaps a lot with Bayswater, Warwick Avenue
and Paddington H&C; a western entrance for Paddington Crossrail would
also be close by.

Sort of. You're right, of course, about other stations being close by, but
neither Bayswater nor Warwick Avenue are stations which offer good routes
to the eastern end of central London (ie the City). Paddington H&C would
of course cease to exist if the H&C was transferred to Crossrail - unless
you're envisaging a very exotic configuration of the subsurface lines!
People could go to Paddington to catch Crossrail, but that involves
pushing local commuters through a busy mainline station. This western
entrance could make that fairly painless, though; i'm afraid i don't know
the details of that.

Nonetheless, there's still the fly-under to think about: it gets to ground
level ~50 m west of Lord Hill's Bridge, which is pretty much level with
the planned Crossrail portal, so there wouldn't be room for a
straightforward junction here; you'd need some sort of weird junction on a
slope thing. Realistically, if you want to use it, you either have to move
the portal east, or you have to leave room for a junction inside the box,
so the branch can run for a hundred metres or so at a bit below ground
level to link up with the fly-under. Or something - IANACivilEngineer.

An alternative would be to ditch it, and add a new fly-under/over around
Westbourne Park, rebuilding the station a little to the east; this would
have the advantage that the station would be upwards of the junction, and
so would be served by more trains. It might also enable some creative use
of the H&C platforms and tracks from Paddington to Westbourne Park by
mainline trains; suggestions on a postcard, please.

Or i suppose you could run the box into the fly-under, run all Crossrail
trains along the existing H&C track on the south side of the formation,
and do something at Westbourne Park to allow them to carry on west - build
a new pair on the south side beyond the junction if possible, or else on
the north side, as is planned, then either provide a fly-under/over to get
there, or put in crossovers to shuffle all the pairs north one position.
Hang on, i'll try some ascii art:

Key:

--- track
+ joining of track
X crossing of track
direction of train
### Westbourne Park platform
.... line heading off to Hammersmith
nU line label; mainlines 1-3 and Crossrail X, plus Up/Down

Now:

1U ------------------------------ 1U
1D ------------------------------ 1D
2U ------------------------------ 2U
2D ------------------------------ 2D
3U ------------------------------ 3U
3D ------------------------------ 3D
/--------------------- XU
### //--------------------- XD
XU ...------//
XD ...------/
###

New pair to the south:

1U ------------------------------ 1U
1D ------------------------------ 1D
2U ------------------------------ 2U
2D ------------------------------ 2D
3U ------------------------------ 3U
3D ------------------------------ 3D
XU -----------+------------------ XU
XD ----------X+---\ ### /-------- XD
XU ...------// \---/
XD ...------/

New pair to the north, with fly-under (flying tracks not shown):

XU -----)
XD ----)
1U ------------------------------ 1U
1D ------------------------------ 1D
2U ------------------------------ 2U
2D ------------------------------ 2D
3U ------------------------------ 3U
3D ------------------------------ 3D
(-+------------- XU
(-X+---\ ### /--- XD
XU ...----------// \---/
XD ...----------/

New pair to the north, with shuffle:

1U -------------\
1D -------------\\
2U ----------+---X+-------------- 1U
2D ----------+X---+-------------- 1D
3U -------+---X+----------------- 2U
3D -------+X---+----------------- 2D
XU ----+---X+-------------------- 3U
XD ----+X---+-------------------- 3D
\\---------+------------- XU
\--------X+---\ ### /--- XD
XU ...----------// \---/
XD ...----------/

I expect that's made things crystal clear to all!

Adding a new pair to the south is clearly best, but probably not possible;
the shuffle would then probably be the easiest, but might leave the lines
in the wrong place for things further down. The fly-under/over would get
everything right, but involves either a big bridge or three small ones -
and does mean that the Crossrail line would cross the formation three
times between Paddington and Heathrow, which is frankly silly.

tom

--
Demolish serious culture!

Tom Anderson wrote:
On Tue, 1 Aug 2006, Dave Arquati wrote:

Tom Anderson wrote:
On Tue, 1 Aug 2006, Dave Arquati wrote:

Bob wrote:

to turn round of half the trains at Paddington which seems a wasted
opportunity - taking over the Hammersmith end of the H&C removes a
busy junction on the Circle Line with knock on effects for that
lines reliability would seem a low cost no brainer.

I agree, but I think ensuring the central tunnel gets built is most
important - adding on too many extra bits here and there could kill
the whole project. Hammersmith could come later...

Except Hammersmith involves noodling about quite near the Paddington
portal (if you want to have a station at Royal Oak, or if you want to
use the existing H&C fly-under, at any rate), so if you don't at
least build in the possibility at the start, you might not be able to
do it later.

I'm not sure it would be worth keeping Royal Oak in this scenario
(gasp). Its catchment area overlaps a lot with Bayswater, Warwick
Avenue and Paddington H&C; a western entrance for Paddington Crossrail
would also be close by.

Sort of. You're right, of course, about other stations being close by,
but neither Bayswater nor Warwick Avenue are stations which offer good
routes to the eastern end of central London (ie the City). Paddington
H&C would of course cease to exist if the H&C was transferred to
Crossrail - unless you're envisaging a very exotic configuration of the
subsurface lines! People could go to Paddington to catch Crossrail, but
that involves pushing local commuters through a busy mainline station.
This western entrance could make that fairly painless, though; i'm
afraid i don't know the details of that.

I don't think there are currently any plans for a western entrance; it
was my suggestion to mitigate a closure of Royal Oak. such an entrance
would be at the junction of Eastbourne Terrace and Bishops Bridge Road.
This would totally separate local travellers from mainline users, and it
wouldn't have to be a grand entrance - Royal Oak isn't particularly
busy, and its passengers would be distributed amongst other stations.

What's wrong with Bayswater for the City? The Circle line has the same
frequency as the H&C, and if Wimblewares were extended eastwards to
compensate for the loss of the H&C, it would see a doubling of frequency.

(The configuration I'm imagining is as now but with no H&C and with
Wimblewares extended from Edgware Road to Whitechapel/Barking... call it
the Wimblebark if you will!)

Paddington H&C could either become extra platforms for mainline trains,
or stabling sidings for the Circle line or Wimbleware(bark).

Nonetheless, there's still the fly-under to think about: it gets to
ground level ~50 m west of Lord Hill's Bridge, which is pretty much
level with the planned Crossrail portal, so there wouldn't be room for a
straightforward junction here; you'd need some sort of weird junction on
a slope thing. Realistically, if you want to use it, you either have to
move the portal east, or you have to leave room for a junction inside
the box, so the branch can run for a hundred metres or so at a bit below
ground level to link up with the fly-under. Or something -
IANACivilEngineer.

An alternative would be to ditch it, and add a new fly-under/over around
Westbourne Park, rebuilding the station a little to the east; this would
have the advantage that the station would be upwards of the junction,
and so would be served by more trains. It might also enable some
creative use of the H&C platforms and tracks from Paddington to
Westbourne Park by mainline trains; suggestions on a postcard, please.

This seems like the best solution, involving the least reconfiguration
of Crossrail. The only problem is whether there's room for grade
separation or not - maybe that could be found through careful
redevelopment of the bus garage.

I'm not sure whether stopping Heathrow/Maidenhead trains at Westbourne
Park would be that useful or not; Ealing Broadway could already be used
for people travelling from the west to Shepherd's Bush or Hammersmith,
and an extra stop on those services might be irritating for travellers
from further out.

(snip pretty ascii art which was useful but needs no comment!)

--
Dave Arquati
Imperial College, SW7
www.alwaystouchout.com - Transport projects in London

Dave Arquati wrote:
What's wrong with Bayswater for the City? The Circle line has the same
frequency as the H&C, and if Wimblewares were extended eastwards to
compensate for the loss of the H&C, it would see a doubling of frequency.

(The configuration I'm imagining is as now but with no H&C and with
Wimblewares extended from Edgware Road to Whitechapel/Barking... call it
the Wimblebark if you will!)

Building on from the H&C Circle line tea cup ideas - how about two
interposed reverse loops. How about existing H&C trains (to and) from
Barking continuing from Edware Road via Bayswater and Victoria to
rejoin the original route at Aldgate East - whilst Wimbleware continues
to and from Liverpool Street, Aldgate and Victoria to join its original
route at Earls Court.

Dave Arquati wrote:

I don't think I'd bother going that far. Taking over as far as
Hammersmith is probably OK because Crossrail serves several of the H&C's
most important destinations. However, it goes nowhere near Victoria and
is a bit of a walk from places like Monument in the City. We know that
Richmonders don't want Crossrail to replace their District service, and
Ealing Broadway will already have Crossrail.

Good points, against which I have no argument.


Shared running would also be a very poor idea as the trains would be so
different and there would be the inevitable performance pollution.

Shared running would only start east of Turnham Green. Interlining
with the North London Line should not be very difficult. It only has
about 3 to 4 trains per hour. The Piccadilly Line is a real problem.
Perhaps the solution to that would be to have CrossRail also take over
the Rayners Lane service. This would double Piccadilly Line service to
Heathrow.

The problem here is the purchase and demolition of all that has been
built in the intervening years. There is also the issue of the many
differing platform lengths en route to the two termini.

However, if there District Line was reduced to an Upminster to
Wimbledon service plus an Edgware Road to Kensington Olympia shuttle,
Circle Line operation would become simplicity itself!

Crossrail operation would be a lot more complicated, though!

Not really given the minimal amount of shared track, i.e. the NLL. I
don't think this idea is do-able, but it beats turning 24 tph back at
Paddington.


Adrian.

wrote:
Dave Arquati wrote:

I don't think I'd bother going that far. Taking over as far as
Hammersmith is probably OK because Crossrail serves several of the H&C's
most important destinations. However, it goes nowhere near Victoria and
is a bit of a walk from places like Monument in the City. We know that
Richmonders don't want Crossrail to replace their District service, and
Ealing Broadway will already have Crossrail.

Good points, against which I have no argument.

Shared running would also be a very poor idea as the trains would be so
different and there would be the inevitable performance pollution.

Shared running would only start east of Turnham Green. Interlining
with the North London Line should not be very difficult. It only has
about 3 to 4 trains per hour. The Piccadilly Line is a real problem.
Perhaps the solution to that would be to have CrossRail also take over
the Rayners Lane service. This would double Piccadilly Line service to
Heathrow.

Not sure exactly what you're suggesting here. Is it:
- Crossrail via Ladbroke Grove & Shepherd's Bush to Rayners Lane via
Ealing Common and to Richmond via Gunnersbury
- District continues to run to Ealing Broadway...?

That results in four western Crossrail branches with shared use between
Gunnersbury and Richmond, between Hammersmith and Turnham Green and
between Acton Town and Ealing Common!

One of the issues raised by the Montague report was that too many
branches at each end would mean poorer reliability, as it's more
difficult to ensure that trains arrive at the core section on time for
their path - making achievement of the 24tph core service difficult.
Having so much shared use to the west would be a recipe for disaster.
The beauty of taking over as far as Hammersmith is that it is
self-contained *and* reduces the number of services trying to interleave
on the northern Circle.

The problem here is the purchase and demolition of all that has been
built in the intervening years. There is also the issue of the many
differing platform lengths en route to the two termini.

However, if there District Line was reduced to an Upminster to
Wimbledon service plus an Edgware Road to Kensington Olympia shuttle,
Circle Line operation would become simplicity itself!
Crossrail operation would be a lot more complicated, though!

Not really given the minimal amount of shared track, i.e. the NLL. I
don't think this idea is do-able, but it beats turning 24 tph back at
Paddington.

I'd say four branches is complicated! Making good use of the 14tph
(24tph really would be a waste!!) seems eminently sensible - but I'd
rather have Crossrail built with those 14tph wasted to begin with, but
scope for future expansion, than see a bloated single-phase project sink.


--
Dave Arquati
Imperial College, SW7
www.alwaystouchout.com - Transport projects in London

On Fri, 4 Aug 2006, Dave Arquati wrote:

wrote:
Dave Arquati wrote:

Shared running would only start east of Turnham Green. Interlining
with the North London Line should not be very difficult. It only has
about 3 to 4 trains per hour. The Piccadilly Line is a real problem.
Perhaps the solution to that would be to have CrossRail also take over
the Rayners Lane service. This would double Piccadilly Line service to
Heathrow.

Not sure exactly what you're suggesting here. Is it:
- Crossrail via Ladbroke Grove & Shepherd's Bush to Rayners Lane via Ealing
Common and to Richmond via Gunnersbury
- District continues to run to Ealing Broadway...?

That results in four western Crossrail branches with shared use between
Gunnersbury and Richmond, between Hammersmith and Turnham Green and between
Acton Town and Ealing Common!

One of the issues raised by the Montague report was that too many
branches at each end would mean poorer reliability, as it's more
difficult to ensure that trains arrive at the core section on time for
their path - making achievement of the 24tph core service difficult.

This is all very true, and i'd agree that the suggested massive takeover
in the west would be a bad idea. However, i really can't believe that
Crossrail can only support *two* branches at each end - three should be
doable, even if not four.

In particular, if there was room for one inward train to wait on each
branch, it should be trivial - divide the hour into 30 2-minute slots,
group these into four, and allocate them one to each branch, plus a spare
(so trains go Shenfield, Dartford, Broxbourne, no train). Set up the
schedule so that each branch delivers a train at the appropriate time. If
a train is delayed, and misses its slot, it sits and waits until the next
empty slot comes round - which could be a spare slot, or a slot missed by
a delayed train on another branch. The worst-case wait would be four
minutes, if a train just missed its slot and had to wait for trains from
both other branches to go through. A slightly better arrangement might be
to group the slots into six five-slot bundles, with four trains and a
spare, so that the position of the spare slot with respect to each branch
changes; otherwise, you get worse delay behaviour on one branch.

In principle, this would work for more branches, but as you increase the
number of branches, the worst-case wait gets worse. Although, if you have
several branches missing slots, the average case might not ...

tom

--
Sapere aude!

Dave Arquati wrote:
wrote:
Dave Arquati wrote:

I don't think I'd bother going that far. Taking over as far as
Hammersmith is probably OK because Crossrail serves several of the H&C's
most important destinations. However, it goes nowhere near Victoria and
is a bit of a walk from places like Monument in the City. We know that
Richmonders don't want Crossrail to replace their District service, and
Ealing Broadway will already have Crossrail.

Good points, against which I have no argument.

Shared running would also be a very poor idea as the trains would be so
different and there would be the inevitable performance pollution.

Shared running would only start east of Turnham Green. Interlining
with the North London Line should not be very difficult. It only has
about 3 to 4 trains per hour. The Piccadilly Line is a real problem.
Perhaps the solution to that would be to have CrossRail also take over
the Rayners Lane service. This would double Piccadilly Line service to
Heathrow.

Not sure exactly what you're suggesting here. Is it:
- Crossrail via Ladbroke Grove & Shepherd's Bush to Rayners Lane via
Ealing Common and to Richmond via Gunnersbury
- District continues to run to Ealing Broadway...?

No, the Piccadilly would retain the 'fast pair' and continue to serve
Heathrow. Richmond, Ealing Broadway and Rayners Lane would be served
by CrossRail by way of Ladbroke Grove and a re-instated link at
Hammersmith. The District Line would become an Upminster to Wimbledon
and Edgware Road to Kensington Olympia Service.

That results in four western Crossrail branches with shared use between
Gunnersbury and Richmond, between Hammersmith and Turnham Green and
between Acton Town and Ealing Common!

The only shared use would be between Gunnersbury and Richmond.
CrossRail would utilize the trackes currently occupied by the District
Line.

One of the issues raised by the Montague report was that too many
branches at each end would mean poorer reliability, as it's more
difficult to ensure that trains arrive at the core section on time for
their path - making achievement of the 24tph core service difficult.
Having so much shared use to the west would be a recipe for disaster.
The beauty of taking over as far as Hammersmith is that it is
self-contained *and* reduces the number of services trying to interleave
on the northern Circle.

And I think that is a fair point. Whilst I think this conversation is
interesting, I don't see the idea I have outlined as a practical
option. The link at Hammersmith pretty much rules it out. But, under
this idea, Circle Line working becomes much simpler. The H&C goes away
and the District is less complex.

The problem here is the purchase and demolition of all that has been
built in the intervening years. There is also the issue of the many
differing platform lengths en route to the two termini.

However, if there District Line was reduced to an Upminster to
Wimbledon service plus an Edgware Road to Kensington Olympia shuttle,
Circle Line operation would become simplicity itself!
Crossrail operation would be a lot more complicated, though!

Not really given the minimal amount of shared track, i.e. the NLL. I
don't think this idea is do-able, but it beats turning 24 tph back at
Paddington.

I'd say four branches is complicated! Making good use of the 14tph
(24tph really would be a waste!!) seems eminently sensible - but I'd
rather have Crossrail built with those 14tph wasted to begin with, but
scope for future expansion, than see a bloated single-phase project sink.

Thank you for correcting my 24 tph! That is what 9 hours COBOL
progamming does to the brain! :-). And I agree, let's see CrossRail
built. We can campaign for more, and better, branches later.

Adrian.

Bob wrote:
Dave Arquati wrote:
What's wrong with Bayswater for the City? The Circle line has the same
frequency as the H&C, and if Wimblewares were extended eastwards to
compensate for the loss of the H&C, it would see a doubling of frequency.

(The configuration I'm imagining is as now but with no H&C and with
Wimblewares extended from Edgware Road to Whitechapel/Barking... call it
the Wimblebark if you will!)

Building on from the H&C Circle line tea cup ideas - how about two
interposed reverse loops. How about existing H&C trains (to and) from
Barking continuing from Edware Road via Bayswater and Victoria to
rejoin the original route at Aldgate East - whilst Wimbleware continues
to and from Liverpool Street, Aldgate and Victoria to join its original
route at Earls Court.

Nice idea, but I think you have too many trains on the Barking and
Wimbledon arms, as well as the southern Circle. For example, you'd have
16tph between Gloucester Road and Tower Hill before you even added any
Districts from Ealing / Richmond.

I imagine you could eliminate all other Wimbledon services, leaving
16tph on the loop service (8tph via Paddington, 8tph via Victoria)

--
Dave Arquati
Imperial College, SW7
www.alwaystouchout.com - Transport projects in London