Increasing Peak Train Capacity

[#Metro]

Hello,

I'm trying to gather thoughts on how to increase peak (and counter peak) train capacity on all lines in a general way to fit more services in as part of a larger idea to get buses feeding train stations. What kinds of things would be relatively easy to do that would allow more passenger throughput during peak hours? I think there was discussion on this forum years ago going though all the options but I can't seem to find it anymore.

Some ideas include: Fewer seats, better signalling (keen to hear more about the potential signalling upgrades), minor upgrades (extra double track, passing track in certain locations, LX removal in key areas etc).

If there are ideas for specific lines (i.e, please post indication which line/section). I am particularly interested in anything that would increase frequency or capacity on the Ferny Grove, Cleveland and Shorncliffe lines.

Thanks! I hope to put all feedback to good use in due course.

[Golliwog]

Faster turn-around times. I posted a while ago (can't remember which thread or when exactly) about how long some of the trains in the AM peak dwell at FG while turning around. I only looked at the FG in isolation but with the existing rolling stock allocated to the FG I think I worked out that with the exception of one service (around 7:40 I think?) that 10 minute frequency was achievable. Obviously there would be changes when you also review the other lines.

Until QR start quarantining particular units to particular lines, I wouldn't support removing seats from units. I'd focus in the medium term on signalling improvements to allow closer services, and smarter timetabling including crew changes to allow more services to be run. The 8:07 ex-FG service is turned around in 1-2 minutes by switching crews. I wouldn't propose such a short changeover for every service, but when you have some units that sit for over 20 minutes, I think that making better use of the trains we have is important. If that means a bit more expense on staffing to provide an extra few crews to allow for the swaps then so be it.

[#Metro]

Thanks Golliwog.

A few questions on details:

1. What is preventing faster turn around times now?

2. How much extra capacity do you think this would provide (i.e. one extra service per hour, increased frequency to X services per hour, X passenger capacity freed up etc)?

3. What signalling improvements would we need (ETCS? ATP? something else)?

I know it is detailed but I need to be sure I have a good handle on what people are suggesting.
Thanks again - anyone else with ideas?

[#Metro]

Derwan, any thoughts / suggestions for the Shorncliffe Line?

[Golliwog]

1. What is preventing faster turn around times now?

Nothing really as far as I'm aware? Someone inside QR would have a better idea on that. But to give you an idea of what is possible, a while ago one of the 6:xx am services from FG failed at the platform and wasn't removed until after AM peak (from memory I got one of the services between 7:30 and 8:00 and it was still there) and the entire AM peak operation went of without any hitches (as far as I'm aware) using just the one platform. I assume there were therefore a few outbound delays as I know of a few services that arrive just before a service leaves or right at the same time, but inbound there didn't seem to be any issues.

2. How much extra capacity do you think this would provide (i.e. one extra service per hour, increased frequency to X services per hour, X passenger capacity freed up etc)?

Not sure, but as I understand it the typical turnaround without a crew swap can be done in 8 minutes or thereabouts so this would allow you to drop below that. Putting a hard number on it though is a bit pointless as service frequency is really going to be limited for a while by rolling stock availability, and this is just a method of reducing the time where a unit is idle. As the FG line is now bottleneck free, the aim should be to operate the best service on the line with the least number of rolling stock as you no longer need to pad the timetable to make sure trains pass at the right locations.

3. What signalling improvements would we need (ETCS? ATP? something else)?

I'm not overly familiar with all the different types of in-cab signalling, but as I understand it the FG line is operating with 3-phase signals currently. I think the end goal should be some level of in-cab, preferably something that could be easily upgraded to a more sophisticated in-cab system when it may eventually be needed down the track. I think the initial goal should be in-cab signalling for Park-Rd to Eagle Junction with the branches continuing with line-side signals until such time as an upgrade is warranted or maintenance becomes an issue.

Although of course the best signalling can't help level crossing incidents or bridge strikes.

[#Metro]

Thanks for this Golliwog.

Not sure, but as I understand it the typical turnaround without a crew swap can be done in 8 minutes or thereabouts so this would allow you to drop below that. Putting a hard number on it though is a bit pointless as service frequency is really going to be limited for a while by rolling stock availability, and this is just a method of reducing the time where a unit is idle. As the FG line is now bottleneck free, the aim should be to operate the best service on the line with the least number of rolling stock as you no longer need to pad the timetable to make sure trains pass at the right locations.

Good. Let's not worry about rollingstock availability at the moment - that can always be solved by buying more trains which is relatively straightforward from a technical perspective (politics may be different though).

I'm not overly familiar with all the different types of in-cab signalling, but as I understand it the FG line is operating with 3-phase signals currently. I think the end goal should be some level of in-cab, preferably something that could be easily upgraded to a more sophisticated in-cab system when it may eventually be needed down the track. I think the initial goal should be in-cab signalling for Park-Rd to Eagle Junction with the branches continuing with line-side signals until such time as an upgrade is warranted or maintenance becomes an issue.

This is very good suggestion, particulary with regards to the strategy of in-cab signalling of the core sections and then leaving the lines further out for later. Do you (or anyone with the technical know-how) think that this approach would also reduce the number of signal faults, particularly on the core section where it is critical?

[petey3801]

Using crew changes at Ferny Grove would certainly speed things up a little bit.

The main reason why trains hang around out there so long at the current time is simply because of the outdated timetable that is still working around no-longer-existant issues like single track. The new timetable I am sure will be rid of this.

8mins is the minimum turnaround time for a 6-car train with one crew, as is set out in the traincrew agreement. This is to allow the driver to properly secure the train on arrival, allow a quick toilet break if needed and reach the other end of the train, set up, complete pre-departure tests and be ready to go. With turnback crews, the toilet break time and walking time is no longer needed, so trains can be turned back in 1-2mins. Main problem that this brings along is reliability. As Golli said, you wouldn't want to do this for every train, as any minor delay (even something like a wheelchair requiring assistance to board/leave the train somewhere along the line) will delay the next run, which isn't good practice. So a buffer of, say, 5mins would be desirable with the current network, to allow for a bit of late running etc.
In the timetable I proposed a couple weeks ago, the way I managed some terminating turnbacks was by use of a turnback crew. Crew A brings the train in to the terminus, crew B sets up and departs in the other direction shortly after. Crew C work the next train in to the platform, by which time crew A have had their short break and are waiting to take that train back on arrival, and so on. It is currently how the 15-min Ferny Grove trains between the peaks are run (so the crew jobcards didn't need to be completely re-written, just a few new ones added).

As for what is needed for extra frequency etc:
Current signalling on the Ferny Grove line is 3-aspect signalling, up to 1km apart (particularly between Bowen Hills and Mitchelton). With the current signalling, a train can't depart (for example) Newmarket until the train in front has completely departed Enoggera.
More signals (500m apart at most) would certainly help this. Signals every 250-300m would be even better. Upgrade to 4-aspect signalling would also help.

A small change to the signalling at Newmarket and Alderly, or complete removal of the level crossings at those stations, would help with signalling also. Currently, all stations trains depart Wilson and Newmarket on a restricted (yellow) signal and approach Newmarket and Alderly respectively on a red. The red clears to proceed as the train stops. This slows trains down considerably due to the speed restrictions placed on trains on restricted signals and approaching reds. While traveling on restricted signals, trains must not proceed at faster than 75% of track (road) speed for the section. Approaching a red signal, trains must be traveling at no more than 40km/h 150m away from the signal (approx. 1 platform length) and no more than 20km/h over the AWS magnet. So, between Wilston and Alderly, trains are a minimum of 25% slower than need be, with much slower station approaches than required if traveling on green signals. This isn't even taking in to account the harsher speed restrictions in place for 160/260 Class units (50% road speed on single yellow, 30km/h max 150m out from red). So, a change in signalling (or removel of LX) at Newmarket and Alderly would speed things up straight away, but would slow things down for road users, as level crossings need to activate earlier.

Shorncliffe line, similar issues re: signalling as is on the Ferny Grove line. Shorter signal spacing with 4-aspect signalling will do a lot to increase headways. Shorncliffe line only has the one LX that creates the same issues as the Ferny Grove line with regards to signalling, and that is the first level crossing (outbound) after Sandgate. This has recently been changed so trains don't actually run on restricted signals anymore, as we arrive in to Sandgate on Green signals, with the starter off Sandgate being yellow and the signal before the LX at red. This all steps up before the trains depart, so in reality, it doesn't effect train running.

Cleveland line, once again, closer signal spacing (would certainly help between Morningside and Norman Park, both directions) and 4-aspect signalling would help greatly. Signal spacing is a little more difficult on this line due to the freight trains running on the line as well, but 4-aspect signalling helps when reducing signal spacing.
Single line outbound of Manly creates a lot of capacity reduction. Selective duplication along this section would help. Birkdale to Cleveland would be relatively easy to do and requires few bridges. Most of the right of way is there already. Lota to Manly would help as well, leaving just Lota to Birkdale as single track with the loop at Thornside. The Lota to Birkdale secion has the most bridges and would be the more expensive section to do, with the Lota to Thornside being the worst due to the river crossings, IMO. Ormiston to Cleveland requires no bridges, only needs a second platform and station upgrade at Ormiston on top of the track works. Ormiston to Wello Point requires one bridge on the WP side of Ormiston across the road and a small bridge along the straight across a walkway/stream, so wouldn't cost much. Wello to Birkdale requires a small bridge across the storm drain and walkway just on the WP side of Birkdale, but also requires work on the road overbridge on the Birkdale side of Wello Point.
A bit of triple track in some places could help with express trains as well as giving freighters another track to run on without interfering with the all stopper suburbans, but this certainly isn't the most pressing issue. Something like Wynnum North to Morningside would be fairly simple as most of the right of way is already there = few resumption costs. But, like I said, that's not urgent at the moment.

[#Metro]

Dear Petey,

Thank you for this very detailed and valuable perspective. It is exactly what I am looking for.

Does anyone have any views on Doomben line services during peak times (there seems to be a handful in peak times both to and from the city). If these were hypothetically removed from peak, which lines would benefit?

I am also thinking of the possibility of Eagle Junction Termination for Doomben Trains... all contributions and ideas (yes, even 'crazy' ones) will be considered!!

[#Metro]

Also, as an aside, if Doomben trains were removed from running to the CBD and train paths were freed up, where could they go? Would they be able to be re-allocated to the Ferny Grove Line or Shorncliffe Lines? Are there restrictions on which lines the paths could be re-allocated to (e.g. tracks, conflicts)?

[BrizCommuter]

http://brizcommuter.blogspot.com.au/2011/02/signalling-case-study-paris-rer.html
This article may be of interest.

[somebody]

Also, as an aside, if Doomben trains were removed from running to the CBD and train paths were freed up, where could they go? Would they be able to be re-allocated to the Ferny Grove Line or Shorncliffe Lines? Are there restrictions on which lines the paths could be re-allocated to (e.g. tracks, conflicts)?

The issue then becomes that you can't turn too many trains around at Shorncliffe.  You could send them to Ferny Grove but these would be surplus to capacity.

But I don't see what is so difficult about providing a reasonable service on the Doomben Line.  A bit of work at Ascot??  Come on.  Combine with an all stopping 4tph Airport line and you have your service for Albion and Woolowin right there.  Which sure beats providing that service with trains that would run as far as Kippa-Ring or Caboolture.

[SurfRail]

Petey - do the Doomben line semaphores affect frequency in any meaningful way?

Ascot second platform should be repaired and put back into use so there is a passing loop in operation, enabling 15 minute peak headways.  Eventually they could do the whole line without a huge amount of difficulty.

[petey3801]

Petey - do the Doomben line semaphores affect frequency in any meaningful way?

Not really, but yes at the same time. For all intents and purposes, at present, the semaphores might as well not be there. They're "fixed" at proceed. So, in that way, they do affect frequency, in that you can't send another train in until the train in front has passed the advance starter at Ascot (it's about 250m after Ascot station, towards Doomben). If, for some reason, they are showing 'Danger', then we have to obey and stop, but that'll only happen if one of the wires has snapped etc. at the moment.

I've done a quick bit of working out, and we can run a 4tph service on the Doomben line in current config, although it is very, very tight. Can also do it so it doesn't use any extra train units as a 30min service as well.
Train 1 travels to Doomben as a 6-car. At Doomben, another crew (either already there or who have traveled spare out there) help to split it in to two 3-car trains. Another train that was already there, but on the other platform, has already split and the front 3-cars departs as soon as the other train has arrived.
14mins later, the front 3-cars departs. Once that 3-car train arrives at EJ, another 6-car departs EJ for DBN 1min later. When it arrives at Doomben, the second 3-car departs (16min behind the first), and the process is repeated. So it makes a 14/16min timetable. ie:

EJ dep 07, 37
DBN arr 14, 44

DBN dep 15, 29, 45, 59
EJ arr 22, 36, 52, 06.

Like I said, very tight, but can be done.

[#Metro]

http://brizcommuter.blogspot.com.au/2011/02/signalling-case-study-paris-rer.html
This article may be of interest.

Thanks BC. This is useful as well re:signalling. Will QR trains be able to stop safely at very short headways? I am not familiar with paris, but some of their trains run on rubber tyres and I would imagine that would affect the safe braking distance. What rollingstock does the RER use?

Any additional ideas to boost FG capacity would be welcome also.

PS: Love the blog, keep up the good work.

[petey3801]

http://brizcommuter.blogspot.com.au/2011/02/signalling-case-study-paris-rer.html
This article may be of interest.

Thanks BC. This is useful as well re:signalling. Will QR trains be able to stop safely at very short headways? I am not familiar with paris, but some of their trains run on rubber tyres and I would imagine that would affect the safe braking distance. What rollingstock does the RER use?

Any additional ideas to boost FG capacity would be welcome also.

PS: Love the blog, keep up the good work.

Yeah, QR trains can stop at fairly short headways, it's just a matter of how short, how fast, what else shares the tracks with pax trains etc. That's where speed limits and signalling work in conjunction with each other.
To give you an idea of our stopping rates on QR units, when running on greens, and where speed limits permit, we are taught to be doing 60km/h at the start of the platform, 40km/h in the middle and come to a stop at the end. 150m. But that is only half service braking, usually. With full service braking, in general (it depends on the particular train, of course.. all trains have slightly different braking characteristics, and these can change depending on what end of the same unit you're driving), we can stop from 60km/h in less than 100m. But, it's not the best idea to be using full service braking all the time, as that means when something unexpected happens, you've got no more braking capacity to stop any quicker. No margin for error = bad idea with this sort of thing.

It also highly depends on operational/policy set speed limits, such as our 75% road speed on restricted signals, 40/20 rule on reds (40km/h 150m out, 20km/h over the AWS magnet) etc. In London, Tube drivers often come in to platforms at speed with red signals. Increases throughput, but also increases the risk of a SPAD, should something go wrong.

[#Metro]

Any ideas for the Cleveland line? Is extending the Ferny Grove extras to terminate at Morningside/Cannon Hill feasible?

[STB]

Any ideas for the Cleveland line? Is extending the Ferny Grove extras to terminate at Morningside/Cannon Hill feasible?

The only turnbacks locations that I know of from my planning days on the Cleveland line is Cannon Hill; Murrarie; Lindum and Manly.  I've heard that the Murrarie terminators may be coming back next year in the new timetables, will wait and see.  Forget trying to up the frequency on the line beyond Manly though, single track restricts it to a 15min frequency as happens in peak, but that doesn't give a path for non revenue trains (driver tuitions, rescue trains etc), which can only happen in the off peak.  Duplicated track is really the only go there, and that's a long way off as we all know.

Busiest stations on the Cleveland line - Morningside, Manly, Birkdale and Cleveland.

[#Metro]

The only turnbacks locations that I know of from my planning days on the Cleveland line is Cannon Hill; Murrarie; Lindum and Manly.  I've heard that the Murrarie terminators may be coming back next year in the new timetables, will wait and see.  Forget trying to up the frequency on the line beyond Manly though, single track restricts it to a 15min frequency as happens in peak, but that doesn't give a path for non revenue trains (driver tuitions, rescue trains etc), which can only happen in the off peak.  Duplicated track is really the only go there, and that's a long way off as we all know.

Busiest stations on the Cleveland line - Morningside, Manly, Birkdale and Cleveland.

Thanks STB. I looked at the QR line diagrams and I can't see a termination siding at Cannon Hill, but I can see two at Murrarie. I think that's the closes t one that does the job!

Not really considering mass infrastructure upgrades, small things are OK, so agree with you re: beyond Manly.

[#Metro]

Shorncliff is good for 15min. The single track section takes 2min, 6min to turn around, 2min back, +3min buffer = 13min, so 2min spare. Worse case, a train running too late you terminate at Sandgate (probably need a cross over to do this). at 15min clockface, Shorncliff will see significant improvement over now. Trains run through to city, all stops.

Where is this single track section?

[minbrisbane]

I can only assume he's referring to the Sandgate - Shorncliffe section.

[Arnz]

I have heard that 15 mins to Murrarie in off-peak has been considered, if not confirmed.  For reasons pointed out already.  Perhaps turn the Doomben trains back there instead of Roma Street? Considering weekday off-peak trains to/from Cleveland continue to Shorncliffe those days.

The fastest turnback IIRC is 4 minutes.  Train #UL01 (5:21am RST to Nambour) morning turns back as train #U986 almost immediately by departing NBR at 7:46am,  though granted this is a 3-car service. 

Maybe Pete or someone else can confirm if service #UL01 picks up a extra crew at CAB to assist with the quick turnaround, whilst the crew of UL01 goes on meal break after arrival at NBR.

[STB]

Shorncliff is good for 15min. The single track section takes 2min, 6min to turn around, 2min back, +3min buffer = 13min, so 2min spare. Worse case, a train running too late you terminate at Sandgate (probably need a cross over to do this). at 15min clockface, Shorncliff will see significant improvement over now. Trains run through to city, all stops.

Where is this single track section?

Before the redevelopment of Sandgate station, the single track section was pretty much Deagon to Shorncliffe with cross overs at Deagon (at least those were the rules when we were planning the Boondall special event services and other non revenue trains).  Otherwise it's simply between Sandgate and Shorncliffe, which IMO there needs to be another crossover built at Sandgate instead of the current two now which can only be used by trains coming from Shorncliffe.

Longer term I do wonder if there is space to stow trains and build a small yard at the end of Shorncliffe station, which might help a bit.

[STB]

Oh, by the way, re: Doomben:  My days at QR, the rules were if a train was occupying the Doomben line, no other train was to use it, with Eagle Junction being the last station before the Doomben line was considered as single track running.  The points at Doomben have to be manually done by station staff with notification if you want an extra train to run on the Doomben line with another train already there, which I think only happened once from what I vaguely remember when the races were on at Ascot.

[Golliwog]

Longer term I do wonder if there is space to stow trains and build a small yard at the end of Shorncliffe station, which might help a bit.

Have a look on Google Maps. I think there is if you took a bit of space from the golf course there and set up some sidings next to Friday Street.

RE: Petey and the signalling details, just out of interest if the end goal is to move to in-cab signalling, do you think there's much to be gained out of spending money on reducing signal spacing/shifting from 3-phase to 4-phase signalling in the short term if it will end up as a stranded investment once you go to in-cab.

[petey3801]

some sort of turn back/3rd platform at Manly if you want 15min to Manly and 30min to CL.

Not required, by what I worked out in another thread. I posted a timetable in (I think) the Sector 2 timetable thread, showing it is possible to have 15min services to Kuraby with 30min to BNH and GC on current infrastructure.

My days at QR, the rules were if a train was occupying the Doomben line, no other train was to use it, with Eagle Junction being the last station before the Doomben line was considered as single track running.  The points at Doomben have to be manually done by station staff with notification if you want an extra train to run on the Doomben line with another train already there, which I think only happened once from what I vaguely remember when the races were on at Ascot.

This is true re: the points at Doomben. Shouldn't be the hardest thing in the world to make them remote controlled points though. Also, there is one train every weekday morning that uses the second platform at Doomben. 1B01, which then sits at Doomben for about an hour while 1B03 arrives and departs from p1.

I looked at the QR line diagrams and I can't see a termination siding at Cannon Hill, but I can see two at Murrarie. I think that's the closes t one that does the job!

Cannon Hill has points on either side of the station and signalling set up to allow for turnbacks from either direction. Murrarie also has points and signalling to enable turnbacks from either direction, but the sidings there are for maintenance vehicles only (not electrified).

RE: Petey and the signalling details, just out of interest if the end goal is to move to in-cab signalling, do you think there's much to be gained out of spending money on reducing signal spacing/shifting from 3-phase to 4-phase signalling in the short term if it will end up as a stranded investment once you go to in-cab.

That's a good point. Only problem is the deaded 'when'. It would have to be weighed up in a cost/benefit type analysis whether to put in more signals or just wait until in-cab signalling is done, but the way things are, I think it'll be many years before we see in-cab signalling, personally. For example, they have been saying ATP was going to be put into the 100 series IMUs for years and years (in fact, AFAIK they were built with wiring etc. ready for ATP, even have hatches on the 8? cab end for the ATP cut out, as well as ATP things inside the 8-cab...), but it still hasn't happened, and the 100 series are, what, 15? years old now...

[HappyTrainGuy]

Longer term I do wonder if there is space to stow trains and build a small yard at the end of Shorncliffe station, which might help a bit.

As far as I know any stabling at Shorncliffe was thrown out in favor of a higher capacity stabling yard at Banyo.

[#Metro]

Not a signalling expert here, what's 4-aspect signalling. Is it yellow/double yellow or something else?

[Golliwog]

Not a signalling expert here, what's 4-aspect signalling. Is it yellow/double yellow or something else?

3-phase is like a traffic light. Green when clear, Red if the next track 'block' is occupied and Yellow if the next track block is empty but the one after that is occupied (i.e. suggesting that the train should prepare to stop at the next signal).

4-phase is similar, but you have the 2nd yellow so you can see if there is a train 2 blocks ahead rather than just one. It allows a driver to slow earlier so that overall you can keep a higher average speed by hopefully avoiding catching up to a red signal and having to stop the train then speed up again.

[STB]

Not a signalling expert here, what's 4-aspect signalling. Is it yellow/double yellow or something else?

3-phase is like a traffic light. Green when clear, Red if the next track 'block' is occupied and Yellow if the next track block is empty but the one after that is occupied (i.e. suggesting that the train should prepare to stop at the next signal).

4-phase is similar, but you have the 2nd yellow so you can see if there is a train 2 blocks ahead rather than just one. It allows a driver to slow earlier so that overall you can keep a higher average speed by hopefully avoiding catching up to a red signal and having to stop the train then speed up again.

I thought it went:

Red
Amber Flash - Next signal at Red (Stop)
Amber - Next signal at Amber Flash (Proceed with caution)
Two Ambers - Next signal at Amber (Proceed with caution)
Green - All clear, go at signed speed.

[STB]

Cleveland line, once again, closer signal spacing (would certainly help between Morningside and Norman Park, both directions) and 4-aspect signalling would help greatly. Signal spacing is a little more difficult on this line due to the freight trains running on the line as well, but 4-aspect signalling helps when reducing signal spacing.
Single line outbound of Manly creates a lot of capacity reduction. Selective duplication along this section would help. Birkdale to Cleveland would be relatively easy to do and requires few bridges. Most of the right of way is there already. Lota to Manly would help as well, leaving just Lota to Birkdale as single track with the loop at Thornside. The Lota to Birkdale secion has the most bridges and would be the more expensive section to do, with the Lota to Thornside being the worst due to the river crossings, IMO. Ormiston to Cleveland requires no bridges, only needs a second platform and station upgrade at Ormiston on top of the track works. Ormiston to Wello Point requires one bridge on the WP side of Ormiston across the road and a small bridge along the straight across a walkway/stream, so wouldn't cost much. Wello to Birkdale requires a small bridge across the storm drain and walkway just on the WP side of Birkdale, but also requires work on the road overbridge on the Birkdale side of Wello Point.
A bit of triple track in some places could help with express trains as well as giving freighters another track to run on without interfering with the all stopper suburbans, but this certainly isn't the most pressing issue. Something like Wynnum North to Morningside would be fairly simple as most of the right of way is already there = few resumption costs. But, like I said, that's not urgent at the moment.

I would've thought it'd be easier to duplicate Manly to Thorneside first, given that there isn't really any major bridging work to be done along there, other than going over the creek.  Ormiston I would've thought would be a bit tricky as it seems that it's a man made rise that the rail and station sits on, which by the way, apparently is causing QR problems with the station sinking according to my contacts and needing ongoing remedial work to prevent damage, hence the big drop between train and platform.  Birkdale to Wellington Point would be another easy spot to duplicate, with Birkdale becoming a split station rather than an island platform, might need to move the ticket office as well to the other side, as I'd assume the city bound trains would depart on that new rail and platform heading inbound.

[somebody]

I thought it went:

Red
Amber Flash - Next signal at Red (Stop)
Amber - Next signal at Amber Flash (Proceed with caution)
Two Ambers - Next signal at Amber (Proceed with caution)
Green - All clear, go at signed speed.

That sounds unlikely.  If the flasher broke you could be shown a less restrictive signal that what was intended.

[somebody]

BL/GC The bottle necks for 4t/hr from each of Kuraby, BL and GC is the lack of 4th track from Salsbury to Kuraby. The 4th track needs to be there for so many km because the track speed through here is so pathetically slow. On the timetable it takes a GC train trailiing a all stopper by 3min at Kuraby to Salsbury to be 3min ahead. This will get you 12t/hr from BL/GC lines.

The 4th track would allow more frequent counter peak and off peak services but wouldn't help peak services.  It also wouldn't help speed unless there was realignment.

[petey3801]

I thought it went:

Red
Amber Flash - Next signal at Red (Stop)
Amber - Next signal at Amber Flash (Proceed with caution)
Two Ambers - Next signal at Amber (Proceed with caution)
Green - All clear, go at signed speed.

That sounds unlikely.  If the flasher broke you could be shown a less restrictive signal that what was intended.

As im on my phone, i'll just respond to a few things now and elaborate later when I get home.
STB is mostly correct in the signalling sequencing. Only thing is that not all signals have flashing yellow. From least restrictive to most restrictive, it goes:
Green (all clear)
Double yellow (where fitted) - caution, next signal single yellow
Single yellow - caution, next signal is at Danger (red)
Flashing yellow (where fitted) - Caution/warning - next signal is at danger, with no overlap beyond the next signal (ie: points set against train directly after the signal, train occupying next block, no margin for error etc) - max speed 40km/h
Red - Danger/stop.

Arnz, forgot to reply earlier sorry. Not sure what happens with UL01, but I imagine either a Caboolture crew goes up spare or a Nambour crew signs on and works it back towards the city while the incoming crew have a meal. Not sure though as I don't work past Caboolture.

[BrizCommuter]

This is true re: the points at Doomben. Shouldn't be the hardest thing in the world to make them remote controlled points though. Also, there is one train every weekday morning that uses the second platform at Doomben. 1B01, which then sits at Doomben for about an hour while 1B03 arrives and departs from p1.

Wow, that must win the award for the most inefficient use of rolling stock!

I thought 30min(ish) shoulder peak turnarounds were bad, and the out of service train that waits outside of Mitchelton for ages in the pm peak, and the Tennyson ghost trains reversing at Corinda, etc, etc.

[BrizCommuter]

Longer term I do wonder if there is space to stow trains and build a small yard at the end of Shorncliffe station, which might help a bit.

I'm sure we will see a dead end terminus with toilets immediately beyond the buffers.

[ozbob]

A long tradition in Queensland!  Gympie 1882 ..

[BrizCommuter]

http://brizcommuter.blogspot.com.au/2011/02/signalling-case-study-paris-rer.html
This article may be of interest.

Thanks BC. This is useful as well re:signalling. Will QR trains be able to stop safely at very short headways? I am not familiar with paris, but some of their trains run on rubber tyres and I would imagine that would affect the safe braking distance. What rollingstock does the RER use?

Any additional ideas to boost FG capacity would be welcome also.

PS: Love the blog, keep up the good work.

The RER doesn't use rubber tyred trains. In fact due to improvements to steel wheel technology there is little if no advantage of using rubber tyres, and all completely new Paris metro lines will use steel wheels in the future. This page will answer your question as to what trains Paris RER uses.
http://en.wikipedia.org/wiki/RER

[STB]

This is true re: the points at Doomben. Shouldn't be the hardest thing in the world to make them remote controlled points though. Also, there is one train every weekday morning that uses the second platform at Doomben. 1B01, which then sits at Doomben for about an hour while 1B03 arrives and departs from p1.

Wow, that must win the award for the most inefficient use of rolling stock!

I thought 30min(ish) shoulder peak turnarounds were bad, and the out of service train that waits outside of Mitchelton for ages in the pm peak, and the Tennyson ghost trains reversing at Corinda, etc, etc.

From memory, one of the Boondall specials (for the Brisbane Entertainment Centre), would sit at the Shorncliffe dead end for about two hours with 1 crew babysitting it.

[Set in train]

From memory, one of the Boondall specials (for the Brisbane Entertainment Centre), would sit at the Shorncliffe dead end for about two hours with 1 crew babysitting it.

How expensive is that?

[petey3801]

This is true re: the points at Doomben. Shouldn't be the hardest thing in the world to make them remote controlled points though. Also, there is one train every weekday morning that uses the second platform at Doomben. 1B01, which then sits at Doomben for about an hour while 1B03 arrives and departs from p1.

Wow, that must win the award for the most inefficient use of rolling stock!

I thought 30min(ish) shoulder peak turnarounds were bad, and the out of service train that waits outside of Mitchelton for ages in the pm peak, and the Tennyson ghost trains reversing at Corinda, etc, etc.

Yeah, then there's always the 30-odd min turnback time at Richlands (which isn't so bad really, as the timetable has allowed for the start of Springfield running without adding more trains/changing the timetable at all), plus the 34 or so min turnaround at Cleveland, 26mins at Beenleigh, 32 at Varsity which increases to 62 in the evening when it's hourly services!