Canning Town

[mrfs42]

Bear with me, because I know next-to-nowt about the DLR.....

I spent a bit of time Friday last at Canning Town waiting for a train to Bank - I'd come off a Jubilee train just as a Bank service rumbled off overhead (if I'd not wanted to go and have a grice of the LED cluster Starter and 'headwall' unit on the WB Jubilee platform I'd probably have been fine).

Anyway, whilst waiting my 20 minutes or so for a train to arrive on platform 4 there were lots of trains coming in and out of platform 3 - some through and some reversing. There was a bit of nifty work by the signalling software with a reverse departing and about 20 seconds later a through train arrived - I was suitably impressed.

Anyway - back to the through and reversing trains: where were they coming from? Were the reversers off the Beckton branch and the through trains Bank - Woolwich Arsenal?

[maxtube]

As far as I know, the reversers are Becktons and the throughs were Bank - King George V/Woolwich.

[astock5000]

Yes, the trainsw that reverse there go to Beckton, and the through trains are the Bank - Woolwich Arsenal service. I don't think any trains terminate at King George V now. In the peaks, some trains run from Woolwich - Blackwall.

[DrOne]

Glad you posted this actually.

I was around Canning Town last thursday or friday evening peak and thought it odd that the wait for a through service was 17mins when I got there. Ended up getting the Jub to Canary Wharf just to be confronted with its struggle with the peak crowds!

[superteacher]

King George V retains full reversing facilities, just in case they're needed.
DLR turnarounds can be very quick, as there is no driver to change ends.

[maxtube]

I actually put King George V because I wasn't sure if Woolwich Arsenal had opened!

[rincew1nd]

Maybe it's the person responsible for teaching me about signalling (6 and 2-5-5 all received in one shift) but I too was mightily impressed with Canning Town the other weekend when I went to have a butchers at Woolwich Arsenal*.

I think there was an event on at ExCel as there was lots of people (both officially and touristy-looking) and it was facinating to see a train pulling into the platform 3 from the East at the same time as one coming around the curve from the West. I think there was only about 10seconds of vacant platform at one point.

If it were conventional signalling there'd be allsorts of approach controls and extended overlaps no-doubt; 40 seconds perhaps between platform occupancy?

*Was it just me/the unit I was on (sat against the windscreen in the northern-most car) or do trains "hunt" a lot when heading south?

[mrfs42]

Jan 27, 2009 22:13:33 GMT rincew1nd said:

Maybe it's the person responsible for teaching me about signalling (6 and 2-5-5 all received in one shift)

I'll have you know that took an awful lot of furtive whispering to achieve! ;D ;D

[rincew1nd]

I know, and I'm sure you know how much I valued that day!

The clock's still where we left it!

[Deleted]

Are there photos of the Canning Town junction works? I'd be interested to know if any works have started in the area.

[Deleted]

Jan 27, 2009 22:13:33 GMT rincew1nd said:

I think there was an event on at ExCel as there was lots of people (both officially and touristy-looking) and it was facinating to see a train pulling into the platform 3 from the East at the same time as one coming around the curve from the West. I think there was only about 10seconds of vacant platform at one point.

If it were conventional signalling there'd be allsorts of approach controls and extended overlaps no-doubt; 40 seconds perhaps between platform occupancy?

I recall on the NY subway seeing two trains (partially) in a platform at the same time - one coming, one going...

[Deleted]

There are some exaggerated claims in this thread!

Typical observed platform re-occupation times:-
DLR - 40secs with approx 55m trains.
Northern Line 3 home signal - 60secs with approx 110m trains.
Central Line speed code ATO/ATP - 60secs with approx 130m trains.
Hong Kong MTR SACEM ATO/ATP - 60secs with approx 180m trains.
LU 5 signal speed controlled approach - theoretical 52secs with approx 110m trains. (Not observed!)
Paris Line 14 - 40secs with approx 90m trains.

So the DLR has an approx 10sec improvement over existing LU signalling taking into account train length. However I have observed better performing moving block systems such as on Paris Line 14 (maybe due to different braking rates?)

As for platform reoccupation times during reversing moves at Canning Town. Well the main benefit of moving block will be the ability of the train to close in to just 50m of a junction set against it. This combined with ATO driving may take around 5-10 secs off platform reoccupation time if you are not taking train length into account.

[mrfs42]

I would take exception to 'exaggerated claims' - I was stood at Canning Town with a stopwatch - the observed minimum interval was 26 seconds. ;D

[tubeprune]

It would be helpful to understand the definition of "platform re-occupation". Do you (mrfs and sk) mean the elapsed time between the rear of the first train clearing the platform track and the front of the second train occupying the track? There are many platforms with more than one track circuit.

[Chris M]

Was that the interval between stationary trains, or between the last part of train 1 leaving the platform and the first part of train 2 entering it?

[mrfs42]

The 26 seconds was from one train moving off to the next train arriving - in other words from the doors fully closed on the departing train to the arriving train coming to a stand. I discounted door opening times.

Looking at it from a traditional signalling point of view, door operating can be discounted.

[DrOne]

When the Stratford Intn'l extension opens which will allow 4 different route combinations at Canning Town, I hope it will be quick & easy to figure out which platform the next train to Beckton or Woolwich will depart from, particularly for passengers coming from the Jubilee or from street level.

[Deleted]

That's an intersting thought! I hope we don't end up playing "musical platforms" as you can do at Camden Town, for example. ;D

[mrfs42]

Wouldn't it be nice to have an old retro lightbox at Canning Town at the top of the escalators to direct passengers!

[Chris M]

Jan 28, 2009 17:45:51 GMT @reganorak said:

That's an intersting thought! I hope we don't end up playing "musical platforms" as you can do at Camden Town, for example. ;D

At least Camden Town isn't as bad as Birmingham New Street!

[Deleted]

Jan 28, 2009 14:00:33 GMT tubeprune said:

It would be helpful to understand the definition of "platform re-occupation". Do you (mrfs and sk) mean the elapsed time between the rear of the first train clearing the platform track and the front of the second train occupying the track? There are many platforms with more than one track circuit.

I'm defining it as wheel start or departing train to wheel stop of arriving train, observed in situations when the second train has caught up with the first and has stopped or slowed for the home signal.

Jan 28, 2009 11:16:29 GMT mrfs42 said:

I would take exception to 'exaggerated claims' - I was stood at Canning Town with a stopwatch - the observed minimum interval was 26 seconds. ;D

Sorry, I was referring to a 10 second quote in this thread!

However, I find your 26sec measurement interesting. In years of using the DLR I have never observed a platform reoccupation time of less than 35secs, the figure usually being around 40secs (or considerably longer at Westferry EB).

[rincew1nd]

Jan 28, 2009 21:05:32 GMT @stephenk said:

Sorry, I was referring to a 10 second quote in this thread!

Sorry, that was me. Ten seconds was probably an exageration, based on MRFS' measurement. It certainly felt quick though!

[thirstquensher]

Nah, I've seen 10-15 sec gaps many times before - not just at Canning Town. Crossharbour, Canary Wharf and even Cyprus/Beckton Park on occasions - especially when a unit hot out of Beckton Depot (entering the network at Gallions Reach), usually Bank- or Lewisham-bound, is chasing a Canning Town train (although that's closer to 25-30 sec).

[mrfs42]

I suppose if you look at it as the time that the platform length is unoccupied, then it will reduce down to about 10 seconds - 8 seconds transit time for each train (doors closing/scarpering)* (arriving/doors opening)* isn't that much really.

*apply/ignore as appropriate.

[tubeprune]

Jan 28, 2009 21:05:32 GMT @stephenk said:

Jan 28, 2009 14:00:33 GMT tubeprune said:

It would be helpful to understand the definition of "platform re-occupation". Do you (mrfs and sk) mean the elapsed time between the rear of the first train clearing the platform track and the front of the second train occupying the track? There are many platforms with more than one track circuit.

I'm defining it as wheel start or departing train to wheel stop of arriving train, observed in situations when the second train has caught up with the first and has stopped or slowed for the home signal.

Jan 28, 2009 11:16:29 GMT mrfs42 said:

I would take exception to 'exaggerated claims' - I was stood at Canning Town with a stopwatch - the observed minimum interval was 26 seconds. ;D

Sorry, I was referring to a 10 second quote in this thread!

However, I find your 26sec measurement interesting. In years of using the DLR I have never observed a platform reoccupation time of less than 35secs, the figure usually being around 40secs (or considerably longer at Westferry EB).

So we are looking at a reduced speed run-in which would give a 90-95s signalled headway - at a guess.

[Deleted]

Jan 30, 2009 18:14:35 GMT tubeprune said:

So we are looking at a reduced speed run-in which would give a 90-95s signalled headway - at a guess.

That would be pretty much right I think. I have observed a 60sec headway, but that was with a reduced speed run in and short dwell time.

The current constraints of flat junctions, reversing capacity, and restrictive signalling around Westferry mean that the DLR has (as far as I know) never has a timetable with more than 29tph on the busiest section (Shadwell to Westferry).