Start again?

[ozbob]

It would be both interesting and potentially very useful to consider a new rail network timetable (suburban and urban) as a theoretical exercise.

Let's assume Corinda to Darra is complete, Darra to Richlands in operation, GC is open to Varsity Lakes, duplication to FG is complete, but nothing to increase CBD capacity track wise, that won't happen if at all until 2016 if we are lucky.

What paths would be optimal?  Are present paths the best?  Eg.  Ipswich to Caboolture, Beenleigh to Ferny Grove and so forth.

Are express services needed on on all lines?  Would a simple capacity increase be facilitated by more passing loops at stations? If so what stations?

Does everything need to run through the CBD?  Where would you put stabling?

Any other ideas?   Comments?  Suggestions?

[Arnz]

It would be both interesting and potentially very useful to consider a new rail network timetable (suburban and urban) as a theoretical exercise.

Let's assume Corinda to Darra is complete, Darra to Richlands in operation, GC is open to Varsity Lakes, duplication to FG is complete, but nothing to increase CBD capacity track wise, that won't happen if at all until 2016 if we are lucky.

What paths would be optimal?  Are present paths the best?  Eg.  Ipswich to Caboolture, Beenleigh to Ferny Grove and so forth.

Are express services needed on on all lines?  Would a simple capacity increase be facilitated by more passing loops at stations? If so what stations?

Does everything need to run through the CBD?  Where would you put stabling?

Any other ideas?   Comments?  Suggestions?

There would be a few changes, but for the most part many paths would remain the same.

- A simple extension of the present-day Shorncliffe-Corinda (and v.v) service to Richlands (including weekend services).  It would extend the 15 min headways to Darra and would introduce 15 min service to the Ipswich Line (as far as Darra) for the first time in the evenings and weekends.

- Improved Ferny Grove Peak service (something like every 10 minutes), unfortunately at the expense of express services, and off-peak will remain the same (unfortunately) despite increased track capacity.

A few changes I would like to see though

- Doomben Line, increased to half-hourly in line with it's partner line, Cleveland Line.  Off-peak patronage may be lacking, but a inner-city line with hourly or worse patronage is one of the reasons why hardly anyone catches it.  6-car services on the Cleveland Line during off-peak is definitely needed as 3-car services are usually PACKED to the brim.

- The Y junction at Landsborough Station, perhaps change that to a turn-back siding. This would enable increased Sunshine Coast off-peak services as far as Landsborough with its current track capacity.  A normal TransLink liveried bus (operated by Sunshine Coast Sunbus) would connect to the Landsborough terminators, and would operate between Landsborough and Nambour on the present-day Railbus route (basically a scaled down railbus operating in TransLink livery).

[stephenk]

It would be both interesting and potentially very useful to consider a new rail network timetable (suburban and urban) as a theoretical exercise.

Let's assume Corinda to Darra is complete, Darra to Richlands in operation, GC is open to Varsity Lakes, duplication to FG is complete, but nothing to increase CBD capacity track wise, that won't happen if at all until 2016 if we are lucky.

What paths would be optimal?  Are present paths the best?  Eg.  Ipswich to Caboolture, Beenleigh to Ferny Grove and so forth.

Are express services needed on on all lines?  Would a simple capacity increase be facilitated by more passing loops at stations? If so what stations?

Does everything need to run through the CBD?  Where would you put stabling?

Any other ideas?   Comments?  Suggestions?

A very big subject!

Off-peak - 15 minute inner suburban frequency, 30 minute outer suburban.
2tph Ferny Grove to Beenleigh
2tph Ferny Grove to Kuraby *
2tph Airport to Gold Coast (exp) *
2tph Airport to Roma Street (exp)
2tph Shorncliffe to Manly
2tph Doomben/Bowen Hills to Cleveland
2tph Petrie to Ipswich/Rosewood
2tph Nambour/Caboolture to Richlands

* It is actually quite tight to run a 15 min off-peak service to Kuraby, and 30 min Gold Coast expresses.
Assuming no extra infrastructure other than that mentioned by Ozbob.

The above would allow for 15min/4tph off-peak service from Ferny Grove to Kuraby, Bowen Hills to Manly, Petrie to Darra, and Airport to Roma Street.

Peak services - this is quite difficult without extra capacity & infrastructure. Something along the lines of:-
Suburban tracks - 20tph from North split between:-
8tph Ferny Grove Line
4tph Shorncliffe Line
4tph Airport Line (Express)
2tph Doomben Line
2tph Bowen Hills start
Suburban tracks - 20tph from South split between:-
8tph Beenleigh Line
4tph Gold Coast Line (express)
8tph Cleveland Line
Main tracks - 15tph from North split between:-
I won't even start trying to work out the Petrie, Caboolture, Nambour mix
Main tracks - 15tph from South split between:-
7.5tph Ipswich Line (express Darra to City except principle stations)
7.5tph Richlands Line (all stations)

In the peaks, line source/destinations would vary as at present, but services would stick to their suburban or main track groupings (e.g ex-Ferny Grove Line service may run to Beenleigh or Cleveland Lines, but not Ipswich or Richlands Lines). This is to avoid conflicting movements, which decrease capacity and reliability.

[stephenk]

- Improved Ferny Grove Peak service (something like every 10 minutes), unfortunately at the expense of express services, and off-peak will remain the same (unfortunately) despite increased track capacity.

After duplication the lines peak service demand will be probably be at least 8tph rather than the 6tph you suggest.
Why no increased off-peak frequency? The Ferny Grove Line can easily justify a 15min off-peak service. Ferny Grove is the busiest suburban station on the whole CityTrain network by quite a margin, and Mitchelton is pretty high up the list too!

- Doomben Line, increased to half-hourly in line with it's partner line, Cleveland Line.

Great idea, just a bit of a problem when you want to get freight trains down the single track Doomben Line.

- The Y junction at Landsborough Station, perhaps change that to a turn-back siding. This would enable increased Sunshine Coast off-peak services as far as Landsborough with its current track capacity.  A normal TransLink liveried bus (operated by Sunshine Coast Sunbus) would connect to the Landsborough terminators, and would operate between Landsborough and Nambour on the present-day Railbus route (basically a scaled down railbus operating in TransLink livery).

Until the Sunshine Coast Line is opened, then increased service as far as Landsborough should be justifiable if it is integrated with connecting buses.

[Arnz]

- Improved Ferny Grove Peak service (something like every 10 minutes), unfortunately at the expense of express services, and off-peak will remain the same (unfortunately) despite increased track capacity.

After duplication the lines peak service demand will be probably be at least 8tph rather than the 6tph you suggest.
Why no increased off-peak frequency? The Ferny Grove Line can easily justify a 15min off-peak service. Ferny Grove is the busiest suburban station on the whole CityTrain network by quite a margin, and Mitchelton is pretty high up the list too!

- Doomben Line, increased to half-hourly in line with it's partner line, Cleveland Line.

Great idea, just a bit of a problem when you want to get freight trains down the single track Doomben Line.

- The Y junction at Landsborough Station, perhaps change that to a turn-back siding. This would enable increased Sunshine Coast off-peak services as far as Landsborough with its current track capacity.  A normal TransLink liveried bus (operated by Sunshine Coast Sunbus) would connect to the Landsborough terminators, and would operate between Landsborough and Nambour on the present-day Railbus route (basically a scaled down railbus operating in TransLink livery).

Until the Sunshine Coast Line is opened, then increased service as far as Landsborough should be justifiable if it is integrated with connecting buses.

The top 2 I have listed weren't my suggestion.  Its my thoughts on what may possibly and should happen (eg the extension of the Corinda service to Richlands), despite evidence pointing otherwise (eg no increase to off-peak on other lines despite evidence pointing otherwise in some areas).

The bottom 2 I listed (The one's I'd like to see) were my suggestions.  I've had more suggestions, but they weren't in my head at the time of the post. 

Also, Landsborough Station is the main interchange/busiest station of the Sunshine Coast Line, not to mention lack of track capacity beyond Landsborough, so suggestions of extra off-peak runs terminating at Landsborough is sensible.

[somebody]

I think arnz is right, there will be very little change other than greater service City-Darra-Richlands.  They will probably add a couple more services as demand forces them to.

Are express services needed on on all lines?  Would a simple capacity increase be facilitated by more passing loops at stations? If so what stations?

Does everything need to run through the CBD?  Where would you put stabling?

Express services are badly needed in the peak and they should also be consistent. On the Ipswich line that would mean running from Darra express to Indooroopilly, express to Toowong, express to Milton.  At least that's what I would think.

I can't think of anywhere where a "passing loop" could work.  Besides, 3min headway signalling isn't up to the job.  What we probably DO need is some quadding between South Bank and Salisbury to allow the GC line services to pass each other without blocking Beenleigh line trains.

Also, running express such that you catch up to the train in front is a little pointless.  The 7:20am and 7:32am from Ipswich do precisely this, and need to wait a minute or so for the all stopper to catch up.  Don't say, "it's only a minute, it's not worth worrying about", it's stupid still.

I can't think of anywhere that trains don't need to run through the city, which means that's the burning issue for our train service.  What we don't need is extra tracks in places that can't use them.  Sandgate-Shorncliffe duplication, Corinda-Darra quad, Darra-Redbank triple, Ferny Grove-Keperra duplication, I am looking at you!

[mch]

Actually QR used a consultant to come up with the current system of services.
Some I don't completely see the sense of, but could be debated for years without agreement from everyone.

Regardless of this, the current track and signalling system can handle 60 trains per hour in each direction during peak.  Theoretically it could handle it all day, but like QR trying to use 97% of the fleet at peak time the pressure on the network could cause problems if it went on too long.
Some minor additional track and signalling work would make this easier to do, but could be handled now with enough units and crews.

A figure of 36 per hour in each direction is more likely achievable.

Break down by lines if we try 20 minute intervals.

Northside
Caboolture ? 3 with 3 from Nambour
Ferny Grove ? 3
Sandgate ? 3
Airport ? 3
Doomben ? 3
Northgate ? 3

Total 21 (This fits)

Southside
Beenleigh ? 3 with 3 from Gold Coast
Cleveland ? 3
Ipswich ? 3 (possible 3 from Grandchester / Rosewood)
Richlands ? 3
Corinda ? 3

Total 21 (This fits)

I may have missed some, but there is still room to move.

Noel Haynes

[ButFli]

20 minute frequencies are no good. Either bring in 15 minute services or stay with half hourly.

[Emmie]

20 minute frequencies are no good.

Why? 

Either bring in 15 minute services or stay with half hourly

On the Shorncliffe line I'd be very happy to move from 2ph to 3ph.  I can't see any problem with 20 minute intervals, and I certainly wouldn't knock them back while I wait for 4ph.

[Derwan]

On the Shorncliffe line I'd be very happy to move from 2ph to 3ph.  I can't see any problem with 20 minute intervals, and I certainly wouldn't knock them back while I wait for 4ph.

I'm with Emmie!

[stephenk]

Regardless of this, the current track and signalling system can handle 60 trains per hour in each direction during peak.

No it can't. The Inner City Rail Capacity Study showed 25tph and 19tph as being the max capacity per direction for the suburban and main tracks respectively. This adds up to 44tph. Realistically, the suburban tracks will struggle to run more than 20tph.  

A figure of 36 per hour in each direction is more likely achievable.

32 is already operated Northbound in the am peak.

20 minute frequencies are no good. Either bring in 15 minute services or stay with half hourly.

15mins is a much more attractive frequency than 20mins. 15mins is seen by many as a turn up and go service. 20mins isn't.  15mins frequency is also the next logical step after a 30mins frequency.

I have studied how easy it is to improve frequencies, and 20mins frequency is a much easier frequency to timetable to more places than a 15min frequency. Unfortunately, as 20mins frequencies are less attractive, and running a mix of 15 and 20mins frequencies would be very messy, I think we should be looking at 15mins frequencies.

[somebody]

No it can't. The Inner City Rail Capacity Study showed 25tph and 19tph as being the max capacity per direction for the suburban and main tracks respectively. This adds up to 44tph. Realistically, the suburban tracks will struggle to run more than 20tph.  

That sounds more reasonable, but what makes the mains lower capacity than the suburbans?  The suburbans are still restricted by the single platform per direction at Fortitude Valley and Bowen Hills.

20 minute frequencies are no good. Either bring in 15 minute services or stay with half hourly.

15mins is a much more attractive frequency than 20mins. 15mins is seen by many as a turn up and go service. 20mins isn't.  15mins frequency is also the next logical step after a 30mins frequency.

I don't really think of 15 mins as turn up and go, 10mins is getting there, but I'm probably a bit more willing to read a timetable than the average person.

I'm with Emmie, in theory.

I have studied how easy it is to improve frequencies, and 20mins frequency is a much easier frequency to timetable to more places than a 15min frequency. Unfortunately, as 20mins frequencies are less attractive, and running a mix of 15 and 20mins frequencies would be very messy, I think we should be looking at 15mins frequencies.

Given that 15mins already applies to Corinda, that could be the knockout blow for 20mins, unless you run 15mins on the mains, and 20mins on the suburbans.

Do we need to improve frequency to Tennyson?  If it can't support an off peak service, does it need a decent frequency in the peak, like at least every 30mins.  Seems such a waste to have a suburban station with such a bad service.

[stephenk]

No it can't. The Inner City Rail Capacity Study showed 25tph and 19tph as being the max capacity per direction for the suburban and main tracks respectively. This adds up to 44tph. Realistically, the suburban tracks will struggle to run more than 20tph.  

That sounds more reasonable, but what makes the mains lower capacity than the suburbans?  The suburbans are still restricted by the single platform per direction at Fortitude Valley and Bowen Hills.

The answer is quite complicated. Capacity is affected by dwell time+platform re-occupation time+operating margin.

The main tracks at Central have a long dwell time, but reasonably good platform re-occupation time (approx 75secs). However due to there being one track per direction, the operating margin has to be timetabled somewhere, and is usually timetabled at Central. This is less of an issue for the suburban tracks as there are two tracks per direction.

Fortitude Valley has short dwell times, so isn't a capacity constraint.

Bowen Hills has long dwell times (for crew changes), and poor(ish) platform re-occupation time (approx 90secs). This affects both tracks fairly equally. Bowen Hills is probably one of the major capacity constraints for the suburban tracks.

Roma Street is an interesting one. Some operating margin is often timetabled here, although the dwell time usually does not have to be too long. However due to the track layouts, the platform re-occupation time is generally poor, particularly for the main tracks (approx 120secs). Things are also made worse by conflicting train movements by trains entering and exiting service crossing the main tracks. This limits the capacity, as you cannot move a train into or out of the main track platforms whilst a train is crossing these tracks to/from the suburban track platforms.

Reversing capacity may be an issue, but don't quote me on this! The suburban tracks have more places where services can be reversed frequently compared to the main tracks.

[somebody]

Thanks stephenk,

That does bring up one question though: Do we need the Dwell time to be in the City?  CityRail (in Sydney) doesn't have it, but just has the dwell times at the start/end of the journey.  I understand Melbourne does it similarly to Brisbane, but they have a jaw dropping number of tracks so presumably capacity to burn.

[Mozz]

From my lay persons perspective the dwell times particularly in peak work best for central as this is where the majority of passengers get on/off the train - quite often in peak times platforms 5 and 6 have trains arriving at the same times and trying to offload 1000 - 1700 people into a small platform area with limited escalator/stair options for them to move through. This situation inherently can take a few minutes.

[Derwan]

As well as allowing for the number of passengers during peak, dwell times in the city provide an opportunity for late services to "catch up" and depart the city on time.

[mch]

My suggestion of 20 minute service was to provide an intermediate improvement to the current half hour.  This is currently achievable with existing signalling and units, but crews would be an issue.

Translink Transit Authority and the government would have to be forth coming with the money for this.

Yes the target should be 15 but this won't be achievable for some time due to unit numbers.

As far as achieving the 21 per hour and even the 28 per hour for a 15 minute service, the numbers are already capable.

Dwell times at Central do not have to be any more than 30 seconds.  General dwell times at other stations are 20 seconds in off peak.
The only reason they are more at peak time is for 2 reasons:
1. Overcrowding of trains increases loading and unloading times as well as clearance problems from the platforms due to design problems of the station at Central.
2. Delaying at Central in peak times helps with on time departures and catch up times from incoming trains is also smoothed out.
The change times at Bowen Hills can be achieved in a 30 second stop as the only reason a minute is allowed is to further help on time running.

Signal spacing in the city area actually allows for less than one minute intervals. As a matter of fact, even in the diesel service days, trains could arrive Central less than one minute apart.  I know that from personal experience.

A proposal for 60 trains per hour each way in peak hour works except for some confrlict issues with services going over the bridge to South Brisbane.  By the way this would required 200 6 car trains as 90% of the fleet this would mean QR would require 223 six car units.  More than double the existing fleet.

The real issue is the will of the public to talk to their local members and let them know that this is a requirement and not just some hope.

Noel Haynes

[stephenk]

Dwell times at Central do not have to be any more than 30 seconds.  General dwell times at other stations are 20 seconds in off peak.
The only reason they are more at peak time is for 2 reasons:
1. Overcrowding of trains increases loading and unloading times as well as clearance problems from the platforms due to design problems of the station at Central.
2. Delaying at Central in peak times helps with on time departures and catch up times from incoming trains is also smoothed out.

I wonder if your stopwatch is slow? The time taken for passenger movements off and onto a busy train at Central is approx 50-70secs. Many metro systems such as London Underground and Tokyo Metro suffer from dwell times approaching 50secs, and that is with more doors per car for increased passenger flow. A 30sec timetabled dwell time during the peaks is currently impossible at Central!

The change times at Bowen Hills can be achieved in a 30 second stop as the only reason a minute is allowed is to further help on time running.

I've observed this, particularly during the evening peak when there are around 5 services, 3 mins apart. Dwell times typically hover around the 45-60sec mark when there is a crew change. This is required for train problems to be communicated to the new crew. London Underground allow for 40sec dwell times for crew changes.

Signal spacing in the city area actually allows for less than one minute intervals. As a matter of fact, even in the diesel service days, trains could arrive Central less than one minute apart.  I know that from personal experience.

Again, I think your stopwatch is slow. I have consistently timed the platform re-occupation time on platform 6 at Central as being 75secs.

A proposal for 60 trains per hour each way in peak hour works except for some confrlict issues with services going over the bridge to South Brisbane. 

That would be 30tph per track. Given the platform re-occupation time on platform 9 at Roma Street is approx 120secs, then 30tph would only be possible if Caboolture bound trains non-stopped Roma Street! There are many parts of the QR CityTrain network that would currently prevent 30tph per track through the core section.

Very few suburban rail systems can run 30tph. The systems that can achieve 30tph all use distance to go or moving block cab signaling which allows for much shorter platform re-occupation times compared to QR's signalling. 

[somebody]

Very few suburban rail systems can run 30tph. The systems that can achieve 30tph all use distance to go or moving block cab signaling which allows for much shorter platform re-occupation times compared to QR's signalling. 

But should we get such a system?  I'd think it would be cheaper than a new river crossing and underground tracks from somewhere like Yeerongpilly to north of Bowen Hills, although it is less benefit.  At least the latter cuts out some slow bits, about 1km of track and could be built for 140km/h running.

[O_128]

We still need the cross river link anything else is a bandaid solution

[stephenk]

Very few suburban rail systems can run 30tph. The systems that can achieve 30tph all use distance to go or moving block cab signaling which allows for much shorter platform re-occupation times compared to QR's signalling. 

But should we get such a system?  I'd think it would be cheaper than a new river crossing and underground tracks from somewhere like Yeerongpilly to north of Bowen Hills, although it is less benefit.  At least the latter cuts out some slow bits, about 1km of track and could be built for 140km/h running.

To get to anywhere near 30tph you need more than new signalling. You would need more doors/car side, more reversing capacity, more decentralised stabling, grade-seperated junctions (at places such as Park Rd), extra tracks, etc, etc.

[mch]

Let me re-clarify some things.
The discussion was really about off peak schedules and so the 30 second platform dwell time quoted and the crew change times were for the off peak period.
This is all distracting from the real discussion.
The off peak time of 20 minutes is achievable and an improvement on the current half hour pattern.

I agree that 15 minutes is what we need, but in the meantime, 20 is better than 30.
It is sometimes a case of take what you can get and at the moment the offering is looking like nothing unless pressure is put where it can make the difference.

Local members will be interested if it means their jobs in about 30 months time.

Noel Haynes

[stephenk]

Let me re-clarify some things.
The discussion was really about off peak schedules and so the 30 second platform dwell time quoted and the crew change times were for the off peak period.

Crew change times would be the same during peak and off-peak.

At off-peak frequencies, there is absolutely no need to restrict dwell times at Central to 30secs. With 20 min frequencies, there would be 6tph using platforms 5 and 6 at Central. This is a headway of 10mins. If you minus 75secs platform re-occupation time and 60secs operating margin, you can still have a dwell time of up to 7mins45secs.  Platforms 1,2,3,and 4 would have 9tph split between track pairs. That is 4.5tph per track, or a headway of 13mins. Thus you could have a dwell time of up to around 10mins. Given that the majority of passengers travel to Central, it makes sense that during the off-peak trains take a long(ish) dwell time there for delay reduction. It is not uncommon for a train running a few minutes late to regain time at Central. The current dwell times at Central are reasonable for all except the impatient.

So I don't quite see your reasoning for 30sec off-peak dwell times at Central.

[mch]

Quote from stephenk August 27, 2009 09:01:24
Ferny Grove is the busiest suburban station on the whole CityTrain network by quite a margin, and Mitchelton is pretty high up the list too.
Sorry, according to QR, Eagle Junction is the busiest station. If this is not correct, someone has misinformed the minister.
Doomben Line:
Great idea, just a bit of a problem when you want to get freight trains down the single track Doomben Line.

Some track and signalling. Duplicating track from Eagle Junction to the Doomben side of Clayfield will make that a lot easier.
Considering the money spent on road works and tunnels, the money required would be small change.

Quote from stephenk September 05, 2009, 07:44:21 PM
No it can't. The Inner City Rail Capacity Study showed 25tph and 19tph as being the max capacity per direction for the suburban and main tracks respectively. This adds up to 44tph. Realistically, the suburban tracks will struggle to run more than 20tph.

I would be interested in knowing who did this study and seeing the types of parameters used.

Quote from stephenk September 06, 2009, 08:32:37 PM
Again, I think your stopwatch is slow. I have consistently timed the platform re-occupation time on platform 6 at Central as being 75secs.
Can you explain what you mean by re-occupation time?  The 75 seconds figure intrigues me.  As it seems that our trains are going backwards.

Quote from stephenk September 07, 2009, 03:53:40 PM
At off-peak frequencies, there is absolutely no need to restrict dwell times at Central to 30secs. With 20 min frequencies, there would be 6tph using platforms 5 and 6 at Central. This is a headway of 10mins.
Why would you assume that only 6 trains would run through 5 and 6 platforms?
Surely a figure of 9 per hour would be more logical as this would take pressure off the 15 that would otherwise have to run through the platforms at Fortitude Valley and Bowen Hills.

[somebody]

1. Sorry, according to QR, Eagle Junction is the busiest station. If this is not correct, someone has misinformed the minister.
2. Some track and signalling. Duplicating track from Eagle Junction to the Doomben side of Clayfield will make that a lot easier.
3. Can you explain what you mean by re-occupation time?  The 75 seconds figure intrigues me.  As it seems that our trains are going backwards.
4. Why would you assume that only 6 trains would run through 5 and 6 platforms?

1: Is there a link to show these numbers?  I heard that Darra & Oxley were pretty high up.
2: But for what benefit?  Will it stimulate any demand?
3: I think he means the time between one train's doors closing and the next train's doors openning
4: All those trains would be going between the Ipswich line, and the Shorncliffe & Caboolture lines.  20min frequency on the last two means 6tph.

HTH

[Arnz]

Between Northgate and Corinda (Soon to be Darra) is currently 4tph during the weekday off-peak and early evenings.  Decreasing the frequency there is a bad idea.

[mch]

1: Is there a link to show these numbers?  I heard that Darra & Oxley were pretty high up.
2: But for what benefit?  Will it stimulate any demand?
I don't know of any link. The information was given by a QR rep at a meeting that the minister had and I was able to attend. Actually quite a few stations are busy. It really doesn't matter in the long run as long as services really meet the need of the travelling public. At this stage they really don't.


3: I think he means the time between one train's doors closing and the next train's doors opening
I assumed this, but my experience at Central is that this time is closer to 50 seconds, so this made me think that I was missing something.

4: All those trains would be going between the Ipswich line, and the Shorncliffe & Caboolture lines.  20min frequency on the last two means 6tph.
In my suggestion, I had 3 from Nambour, 3 from Caboolture and 3 from Northgate. Now depending on whether these were going to Corinda or whether the Shorncliffe were going to Corinda, these would normally travel via 5 and return through 6.

[somebody]

Between Northgate and Corinda (Soon to be Darra) is currently 4tph during the weekday off-peak and early evenings.  Decreasing the frequency there is a bad idea.

Early evenings?  It finishes between Milton & Corinda around 7pm.  But yes, a decrease in frequency would be ridiculous as these trains have been quite busy when I've been on them.

[stephenk]

Quote from stephenk August 27, 2009 09:01:24
Ferny Grove is the busiest suburban station on the whole CityTrain network by quite a margin, and Mitchelton is pretty high up the list too.
Sorry, according to QR, Eagle Junction is the busiest station. If this is not correct, someone has misinformed the minister.

Passenger movements per day 2009:
Ferny Grove 3783
Eagle Junction 3132

Doomben Line:
Great idea, just a bit of a problem when you want to get freight trains down the single track Doomben Line.

Some track and signalling. Duplicating track from Eagle Junction to the Doomben side of Clayfield will make that a lot easier.
Considering the money spent on road works and tunnels, the money required would be small change.

I would agree, although it's not top of the list of required infrastructure.

Quote from stephenk September 05, 2009, 07:44:21 PM
No it can't. The Inner City Rail Capacity Study showed 25tph and 19tph as being the max capacity per direction for the suburban and main tracks respectively. This adds up to 44tph. Realistically, the suburban tracks will struggle to run more than 20tph.

I would be interested in knowing who did this study and seeing the types of parameters used.

The ICRCS is very well written, and the figures quoted are assuming improvements in infrastructure that need to occur by 2015. I have also seen "official" figures quoting 20tph and 15tph as the max capacity for the main and suburban tracks respectively, which seems more realistic to the current infrastructure.

Quote from stephenk September 06, 2009, 08:32:37 PM
Again, I think your stopwatch is slow. I have consistently timed the platform re-occupation time on platform 6 at Central as being 75secs.
Can you explain what you mean by re-occupation time?  The 75 seconds figure intrigues me.  As it seems that our trains are going backwards.

Platform re-occupation time is time taken from wheel start of the departing train to wheel stop of the arriving train. It is also known as dynamic headway, close in time, run out run in time, roll out roll in time (RORIT), and probably plenty of other names.

Quote from stephenk September 07, 2009, 03:53:40 PM
At off-peak frequencies, there is absolutely no need to restrict dwell times at Central to 30secs. With 20 min frequencies, there would be 6tph using platforms 5 and 6 at Central. This is a headway of 10mins.
Why would you assume that only 6 trains would run through 5 and 6 platforms?
Surely a figure of 9 per hour would be more logical as this would take pressure off the 15 that would otherwise have to run through the platforms at Fortitude Valley and Bowen Hills.

I was assuming:-
3tph Shorncliffe-Richlands and 3tph Caboolture-Ipswich on mains.
3tph Ferny Grove-Beenleigh, 3tph Doomben-Cleveland, and 3tph Airport-Gold Coast on suburbans.

[somebody]

A very big subject!

Off-peak - 15 minute inner suburban frequency, 30 minute outer suburban.
2tph Ferny Grove to Beenleigh
2tph Ferny Grove to Kuraby *
2tph Airport to Gold Coast (exp) *
...
* It is actually quite tight to run a 15 min off-peak service to Kuraby, and 30 min Gold Coast expresses.
Assuming no extra infrastructure other than that mentioned by Ozbob.

The above would allow for 15min/4tph off-peak service from Ferny Grove to Kuraby, Bowen Hills to Manly, Petrie to Darra, and Airport to Roma Street.

Peak services - this is quite difficult without extra capacity & infrastructure. Something along the lines of:-
Suburban tracks - 20tph from North split between:-
8tph Ferny Grove Line
4tph Shorncliffe Line
4tph Airport Line (Express)
2tph Doomben Line
2tph Bowen Hills start
Suburban tracks - 20tph from South split between:-
8tph Beenleigh Line
4tph Gold Coast Line (express)
8tph Cleveland Line
Main tracks - 15tph from North split between:-
I won't even start trying to work out the Petrie, Caboolture, Nambour mix
Main tracks - 15tph from South split between:-
7.5tph Ipswich Line (express Darra to City except principle stations)
7.5tph Richlands Line (all stations)

In the peaks, line source/destinations would vary as at present, but services would stick to their suburban or main track groupings (e.g ex-Ferny Grove Line service may run to Beenleigh or Cleveland Lines, but not Ipswich or Richlands Lines). This is to avoid conflicting movements, which decrease capacity and reliability.

In fact, I am wondering how the above will fit.  Even 2tph Gold Coast and 4tph Kuraby-City would require some sort of passing move in both directions.  Where can that be done?

8tph on the Cleveland line?  Most of those trains would need to terminate by Manly in the outbound direction.
8tph to Ferny Grove?  I that's way to tight on the single track.  6tph would need to be the maximum.  You could run 6tph, with at least some expresses though.

Regarding the Ipswich/Richlands line, I would be thinking something like 6tph off peak broken down like this:
2tph Ipswich-City , stopping all to Oxley, express to Indooroopilly, express to Toowong, express to Milton
2tph Corinda-City , all stops
2tph Richlands-City. all stops
The express trains save about 5 minutes, giving a 15 minute frequency from Darra to the city, and roughly a 10 minute frequency from Toowong/Indooroopilly.  The less popular stations from Corinda in would have to deal with up to a 20 minute wait.

Doing it that way, you could 3tph to Caboolture and Shorncliffe off peak.

[O_128]

most people would be happy 2tph to manly and 2pth to cleveland. it will require antoehr platform at manly though

[stephenk]

most people would be happy 2tph to manly and 2pth to cleveland. it will require antoehr platform at manly though

4tph to Manly, with 2tph of those sent onto Cleveland would not require an extra platform at Manly during the off-peak.

However if the peak frequencies are to be increased, then an extra platform at Manly would be very beneficial. An extra platform would allow a train to have it's terminus dwell without obstructing the inbound or outbound tracks. This would significantly increase the flexibility of the timetable.

[somebody]

Two things that they really need to do differently from at present are:
(1) Predictability.  If a service comes at :17 and :47 it needs to always come at those times.  It's no good having a few at :16 in off peak times, as that can cause people who are running late to miss the train and be justifyably p%ssed  off.  This is the main problem with the service cuts Sun-Thur evenings and Sun mornings
(2) Bunching.  Having a large gap followed by two services in reasonably quick succession isn't good.  Currently, there's a 7:16am train from Ipswich followed by a 7:20am train and a 7:24am train.  There's an 11 minute gap before the 7:16 train and an 8 minute gap after the 7:24am train, with a 25 minute gap to the following train. The 7:20am train expresses through 3 stations, I think, but it's not quite enough to get in front of the Corinda starter and then has to wait.  If this service expressed through a few more stations and the Corinda train left slightly later on a different track, you could at least achieve something.  The 7:16am train comes from Rosewood so could also justifyably skip a few more stations. 

The busses have both of these problems too when you look at multiple routes on the same corridor.

[somebody]

Where would you put stabling?

This is a real problem for some lines, notably Ferny Grove, but I'm not sure if we are using our existing stabling to it's utmost.  Having Mayne stabling and empty or positioning moves inbound around midnight and outbound before the first service isn't a good thing.

[O_128]

Where would you put stabling?

This is a real problem for some lines, notably Ferny Grove, but I'm not sure if we are using our existing stabling to it's utmost.  Having Mayne stabling and empty or positioning moves inbound around midnight and outbound before the first service isn't a good thing.

isnt there room on the doomben line?

[somebody]

Where would you put stabling?

This is a real problem for some lines, notably Ferny Grove, but I'm not sure if we are using our existing stabling to it's utmost.  Having Mayne stabling and empty or positioning moves inbound around midnight and outbound before the first service isn't a good thing.

isnt there room on the doomben line?

Probably.  I wouldn't think it would be too hard to enough stabling for that line.

[brad C]

Stabling on the Doomben line has problems.
Firstly, the electrified track traverses through mainly residential areas and some of the reasons why there are a lot more returns to Mayne of an evening is the effect on residents from stabling activities.
Secondly movements from the Doomben line cannot be routed to either the northern suburbs or the FG line as there are no reverse cross overs. It would be counter productive to run these sets empty through the city then back out to the northern suburbs.

In the 1947 electrification grand plan, large stabling yards were to be located at Bindah, Taigum and Walkuraka, in addition to Mayne. Mind you this scheme was not planned to cover the wider areas of today's network but the foresight was there.

What about the spare electrified sidings at Normanby??

One last problem with remote stabling is staff availability and access.
This is the positive aspect of large stabling facilities located in central areas (Mayne) or satellite cities (Ipswich, Beenleigh/Robina and Caboolture/Nambour).

[Emmie]

In the 1947 electrification grand plan, large stabling yards were to be located at Bindah, Taigum and Walkuraka, in addition to Mayne

What's to stop going back to this now and using Bindha?  Huge amount of land with very little evidence of activity - some could easily be used for stabling, I would think.

And where IS Walkuraka??

[ozbob]

http://en.wikipedia.org/wiki/Wulkuraka_railway_station,_Queensland

Past Ipswich on the way to Rosewood. Plenty of room there.

[O_128]

In the 1947 electrification grand plan, large stabling yards were to be located at Bindah, Taigum and Walkuraka, in addition to Mayne

What's to stop going back to this now and using Bindha?  Huge amount of land with very little evidence of activity - some could easily be used for stabling, I would think.

And where IS Walkuraka??

definently i think over the next 20 years we will see a decentralization of stabling. This will be because of bowen hills becoming practically another cbd and people will not want the noise there.