How far down the Beenleigh line is 15 min off peak frequency possible?

[somebody]

Personally I think the best you'll get with the Beenleigh line in the short to medium term is half hourly services to Beenleigh and the Gold Coast as you currently get overlapped with all stoppers to Kuraby.  Half hourly services to the Gold Coast ain't really that bad.

No, but it is pretty bad on the inner parts of the Beenleigh line.

[colinw]

Indeed.  Half hourly at Park Road or Dutton Park is laughable.

Remember also that there are electrified 3rd lines at Kingston (no platform) & Bethainia - it may be possible to work out some overtaking moves at those locations?

It is a pity Kuraby wasn't made a 4 platform station.  The crossover arrangements there seem to lead to conflicts no matter what you do.

Maybe 20 minutes is the best interim solution?  Although that would lead to all sorts of complexity in the CBD if other lines stayed on a half hourly or 15 minute pattern, as the two patterns would fall into and out of phase with each other.

[somebody]

20 minutes would only work out if it was 20 minutes to the Gold Coast as well.  I really think that the Beenleigh line frequency would need to be an integer multiple of the Gold Coast line frequency.

One crossover somewhere like Altandi from the southbound track to the middle track and some bi-di signaling on the middle track would pretty much solve all the constraints with getting a 15 minute frequency to Kuraby, which also works out reasonably well from a timetabling point of view.  Perhaps add an extra crossover immediately to the north of my previously proposed crossover so the northbound stopping trains from Beenleigh can get out of the way of the Coasties

[colinw]

Eh?  I'm fairly sure all 3 tracks from Salisbury to Kuraby are bidirectionally signalled - I go along Beenleigh Rd quite frequently and have noticed that there are signals above all 3 tracks in both directions.

My comment about 20 vs 15 / 30 frequency was more concerning the effects from Park Road through the CBD if Cleveland remained on a 30 minute frequency.

[somebody]

Eh?  I'm fairly sure all 3 tracks from Salisbury to Kuraby are bidirectionally signalled - I go along Beenleigh Rd quite frequently and have noticed that there are signals above all 3 tracks in both directions.

I was going to say: you still need the crossover(s) for the Kuraby train to get out of the way of the southbound coastie, however, perhaps you don't.

Current running times have the Coastie heading south (from Central) at :24 & :54 and passing Kuraby at (estimated) :53 & :23.  The Beenleigh train heads south at :01 & :31 and stops at Kuraby at :42 & :12.  Adding a Kuraby train at :16 & :46 would reach Kuraby at :57 & :27 which would not get out of the way of the coastie in time.  But perhaps the Beenleigh & Kuraby trains could be timed 3 minutes earlier?  Seems too tight, and the southbound coastie would have to lose a couple of minutes, which is something I am not in favour of.

I want the extra crossover to allow the southbound Kuraby train to use the middle track for the last few stations.

Going as far a Kuraby is well worth while IMO.

[colinw]

Ah, now I follow you - must be having a slow day. Yeah, looking at the diagrams here it seems the only crossovers from the up southern main to the down southern main are at Salisbury (12.667km) and Kuraby (21.234km).  This makes the bidirectional signalling somewhat useless for the purposes being discussed here.

It seems to me that a set of crossovers at around the 16km mark between Banoon & Sunnybank would neatly divide that long section.

cheers,
Colin

[somebody]

I was thinking it should be further south than that, but then, what for unless there is a corresponding crossover from the third road to its north?

By adding a single crossover, we could have 15 minute frequency as far as Kuraby with no impact on the Gold Coast line.  Seems like a good idea to me.  Counter peak may cause some impact on the Gold Coast line though, but that could be virtually eliminated by a second crossover.

Perhaps a suitable subject for a media release.

[somebody]

What happens when/if the triplication extends to Kingston?

I think that might make david's proposal a lot more doable.  Then what would be the ideal location for the crossover(s)?  This is important because I do not think we should be suggesting something which would be a stranded investment in several years time.

I think Altandi is about right, directly after the station in both directions but it would not be possible to run Altandi-Kingston-Altandi and have the crew change ends in 30 minutes.  Therefore, once the triplication extends, we may as well go for david's plan in my view.  I do not think it is doable with adequate reliability right now.

What about impacts post CRR1?  Depends on if there is a quad down to Salisbury or beyond.

[colinw]

Isn't there a study suggesting quad to Banoon as necessary?  It also suggested that any quad down that way would have an Up Up Down Down arrangement, rather than the Up Down Up Down of the Corinda quad.

Your question about the triple extending to Kingston is a valid one, particularly considering the boundary of UrbanLink services is proposed to be Loganlea.

Altandi to Kingston running time in the current timetable is precisely 15 minutes, so Altandi - Kingston - Altandi in 30 is impossible.

[somebody]

Isn't there a study suggesting quad to Banoon as necessary.  It also suggested that any quad down that way would have an Up Up Down Down arrangement, rather than the Up Down Up Down of the Corinda quad.

You are probably thinking of the ICRCS there, which suggested a Fairfied-Banoon quad.  However, with the Yeerongpilly portal, that seems a bit outdated.

God forbid up-down-up-down.

[STB]

Years ago I was listening to some long term planners at TL and they did mention that quadding the Beenleigh line would need to happen eventually.

I think in reality to achieve a 15min frequency on both the Beenleigh lines and Gold Coast lines, they both need their own dedicated tracks (quadding) plus the Cross River Rail.

[somebody]

Years ago I was listening to some long term planners at TL and they did mention that quadding the Beenleigh line would need to happen eventually.

I think in reality to achieve a 15min frequency on both the Beenleigh lines and Gold Coast lines, they both need their own dedicated tracks (quadding) plus the Cross River Rail.

I do not see this happening, nor do I see it as a fair ask, to be frank.

[colinw]

Four tracks to accommodate a mere 8tph in each direction seems like typical Queensland infrastructure fetishism to me.

Our discussion should be guided by the IMMEDIATE need for a 15 minute service to a point as far south on the Beenleigh line as possible.  The question was "how far can 4tph go NOW", not "how many tracks can we build in 10 years time".  If we can sling in a crossover or two and get a viable 4tph service to Kuraby or Kingston now, then that should be enough for the next few years.

The ultimate infrastructure need will be dictated by the 2031 vision of UrbanLink to Loganlea, ExpressLink to Ormeau, CoastLink to Coolangatta plus possibly an UrbanLink service on the Gold Coast as well.

[somebody]

Four tracks to accommodate a mere 8tph in each direction seems like typical Queensland infrastructure fetishism to me.

Yes, but to be fair more services in peak would be operated.  And there needs to be quad at the Yeerongpilly end to allow for overtaking moves.

To be frank, I cringed when I saw RailBoT put forward this plan in its current form.  Limitations are:

• Necessary crossovers not included
• Flow on effects north of the CBD or south of Beenleigh not considered
• Cannot be maintained counter peak
• Likely to have a large and negative effect on reliability
• Peak flow not properly considered either

I'm sorry, I hope that's not too harsh.

[#Metro]

Yes, but to be fair more services in peak would be operated.  And there needs to be quad at the Yeerongpilly end to allow for overtaking moves.

To be frank, I cringed when I saw RailBoT put forward this plan in its current form.  Limitations are:

    * Necessary crossovers not included
    * Flow on effects north of the CBD or south of Beenleigh not considered
    * Cannot be maintained counter peak
    * Likely to have a large and negative effect on reliability
    * Peak flow not properly considered either


I'm sorry, I hope that's not too harsh.

Not at all, its a valid point you raised somebody. The Gold Coast has half a million people AIUI, which in population terms is half way and gaining on a city like Adelaide (1 million)! To think that 2 trains/hour to the GC and 2 trains/hour to Beenleigh is going to hold up over the next 20-30 years is dreaming! There had better not be a sudden increase in fuel costs (carbon tax anyone?) or the rail system won't cope.

Our discussion should be guided by the IMMEDIATE need for a 15 minute service to a point as far south on the Beenleigh line as possible.  The question was "how far can 4tph go NOW", not "how many tracks can we build in 10 years time".  If we can sling in a crossover or two and get a viable 4tph service to Kuraby or Kingston now, then that should be enough for the next few years.

Spot on. Its what can be done now with the minimum of, or or no, infrastructure upgrades.

[colinw]

I am wholeheartedly sick of grandiose sounding big rail projects that deliver nothing at all.  Case in point - Caboolture to Beerburrum.  18 months later and where's the faster or more frequent services?  Timings to Nambour - unaltered.  Tilt timing to Rocky, unaltered.  Trains just sit at Beerburrum using up the slack.  What the heck was all that money spent for then?

Ditto, Salisbury to Kuraby triplication + Ormeau - Coomera and Helensvale - Robina duplications.  Sod all change to the basic service frequencies, just a bit of reliability.  Talk about doing less with more.

How about a bit more focus on doing the most with what we have, instead of the current Queensland pattern of keeping services at the current pathetic level, promising huge rail projects in 5, 10, 15 or even 20 years, and even then failing to deliver much in the way of services once they are built.

[somebody]

Ormeau - Coomera and Helensvale - Robina duplications.  Sod all change to the basic service frequencies, just a bit of reliability.  

Is that right?  Was still half hourly even without the duplications.

Current timetable has a cross between Nerang and Helensvale.  This would have needed to be done at a station with 2 platforms

How did they do the 15 minute peak frequency, or didn't they?

[#Metro]

How about a bit more focus on doing the most with what we have, instead of the current Queensland pattern of keeping services at the current pathetic level, promising huge rail projects in 5, 10, 15 or even 20 years, and even then failing to deliver much in the way of services once they are built.

I agree. Everything under the sun- including shoosh carriages and rottweilers on trains has been trotted out to look like something is happening. There really is one big issue that gets to me: wasting 1/2 hour, or a whole hour per day wasted sitting at a rail station platform!

A lot of money is being spent on hugely expensive projects, but does the service frequencies follow? Even worse is the idea that projects can be started and stopped at will because 'it will be cheaper in the future'. This is leading to an infrastructure snowball- where you have to go back and do projects twice, using up more time and more money and the construction materials go up and up and up all the while having to deal with new problems.

It always costs more to go back and do a project twice.

[colinw]

Ormeau - Coomera and Helensvale - Robina duplications.  Sod all change to the basic service frequencies, just a bit of reliability.  

Is that right?  Was still half hourly even without the duplications.

Current timetable has a cross between Nerang and Helensvale.  This would have needed to be done at a station with 2 platforms

How did they do the 15 minute peak frequency, or didn't they?

It was half hourly offpeak right through to Robina even with single track beyond Ormeau.  I think the cross was at Helensvale during the offpeak period, with some crosses at Coomera in the peak - I can remember sitting at Helensvale for about 4 minutes on a Robina bound offpeak train.  I think there also may have been some hourly frequency in the middle of the day for a while after the line was first (re)opened in the mid 1990s.

I'm not sure what the peak frequencies were before the duplications, from memory one or two extra peak services were added after the duplication, but not many.

[#Metro]

It was half hourly offpeak right through to Robina even with single track beyond Ormeau.  I think the cross was at Helensvale during the offpeak period, with some crosses at Coomera in the peak - I can remember sitting at Helensvale for about 4 minutes on a Robina bound offpeak train.  I think there also may have been some hourly frequency in the middle of the day for a while after the line was first (re)opened in the mid 1990s.

I'm not sure what the peak frequencies were before the duplications, from memory one or two extra peak services were added after the duplication, but not many.

Colinw, this is what I don't understand. Is there any proof that half-baking an infrastructure project is cheaper and better than full-bake? Consider this: every passenger using that line must endure a 4 minute wait. And again on the way home-another 4 minutes, and then multiply by 52 weeks x 5 days x the value of time x 800 people on the train or so

Is it really "cheaper"? How much would it cost to go another 3 km of double track? 100 million?

[colinw]

The "half baking" of the Gold Coast line is a classic case of penny wise, pound foolish.  Instead of building a fit-for-purpose double track line the whole way, then enjoying the benefits, we have been paying for piecemeal extensions & duplications every few years ever since the thing opened.

The Queensland way:

1960: Nerang to Tweed Heads closed
1964: Beenleigh to Nerang & Southport closed.
1980s: new line to Gold Coast promised
1991-2: duplication from Kuraby to Beenleigh, Gold Coast line surveying starts.
1994-5: construction to Helensvale.
1996: Phase 1 opens, double track to Ormeau, single track beyond.  Construction beyond Helensvale continues
1998: Phase 2 - Helensvale to Nerang opens in early '98
1998: Phase 2 complete: Nerang to Robina opens in late '98
2005: Ormeau to Coomera duplication work begins
2006: Ormeau to Coomera duplication complete.
2006: Helensvale to Robina duplication begins
2007: Robina to Varsity Lakes extension begins
2008: Helensvale to Robina duplication complete.
2009: Robina to Varisty Lakes extension complete

End result.  Half hourly basic service pattern, single track remains Coomera to Helensvale.  Extensions to Elanora & Coolangatta?  Who knows when.  Service speed remains hampered by 19th century alignment from Beenleigh to the CBD & conflicts with the all stations Beenleigh service.

The WA way:

1960s: Original line to Rockingham closed.
1999: Legislation for Mandurah line passed.  Some delays occur as a result of a change of Government.  The spur to Thornlie is a legacy of this (the original Mandurah line proposal was to go via Thornlie).
2004: Construction commences
2007: Construction complete

End result: double track the whole way to the main CBD station, including new city underground stations.   Basic service frequency is 15 minutes all day, and as little as 4 minutes in peak.  No conflicts with other services & completely segregated from freight as well.  Nothing much will need to be done to this line for decades, unless the proposed extension to Bunbury goes ahead.

The Mandurah line was criticised in some circles (e.g. the anti-rail West Australian newspaper) for the big upfront cost, but out of Gold Coast & Mandurah which looks like better value now?  My guess is that the projects required to bring the Gold Coast line up to a Mandurah standard service will cost as much or more than building the entire Mandurah line did.

Link to Mandurah timetable..  Gold Coast & Brisbane residents can read it and only dream of the day we have something as good.  The trains operating it are basically the same as our 160 class IMUs.

cheers,
Colin

[#Metro]

The WA way:

1960s: Original line to Rockingham closed.
1999: Legislation for Mandurah line passed.  Some delays occur as a result of a change of Government.  The spur to Thornlie is a legacy of this (the original Mandurah line proposal was to go via Thornlie).
2004: Construction commences
2007: Construction complete

End result: double track the whole way to the main CBD station, including new city underground stations.   Basic service frequency is 15 minutes all day, and as little as 4 minutes in peak.  No conflicts with other services & compeltely segregated from freight as well.  Nothing much will need to be done to this line for decades, unless the proposed extension to Bunbury goes ahead.

The Mandurah line was criticised in some circles (e.g. the anti-rail West Australian newspaper) for the big upfront cost, but out of Gold Coast & Mandura which looks like better value now?

Actually, I think Mandurah was done on the cheap as well. The construction costs per kilometre are unbelievably small.

The Gold Coast is a city of half a million.

http://www.goldcoast.qld.gov.au/t_standard2.aspx?pid=255

The Gold Coast's current population of 515,157* is expected to continue to grow by 13,000 to 16,000 people per year, so that by 2026 Gold Coast City will be home to over 730,000 residents.

And what is at Mandurah huh? Nothing much!!!

[#Metro]

I forgot to mention: The Mandurah line terminates at Mandurah, which is 70.1 km from Perth CBD- much the same distance to Nerang/Robina!!!

[Stillwater]

Don't forget the large population growth over the border in NSW.  It will never happen because of the jurisdictional differences, but a visionary tripartite (federal, Qld and NSW)approach to public transport should be capable of exploring options for extending narrow gauge rail to Byron Bay.  Meanwhile, community groups push for a Newcastle type suburban rail service on the disused Casino-Lismore-Murwillumbah standard gauge line.

[#Metro]

It makes me wonder how a quarter of Coolangatta airport hangs over the border in to NSW.
How did they ever agree on that one?

[longboi]

That's because airports (at least major airports) are on federal land 

[#Metro]

That's because airports (at least major airports) are on federal land

Really! Maybe in the future "major railways" could be added to that list!

[Golliwog]

They can eventually agree on things though. IIRC the alignment for the railway to Coolangatta airport has the line looping around the south of it, through NSW then back into QLD for the station at the terminal.

[#Metro]

Service speed remains hampered by 19th century alignment from Beenleigh to the CBD & conflicts with the all stations Beenleigh service.

I agree. The stations are also ridiculously closely spaced- about 1km apart, which makes the service unbearably tedious. There is a non-infrastructure solution for this, but it is a political hot potato. If frequencies were increased beyond the 15 minute timetable, skip stop operation would be possible. At the moment, buses are faster in parts- which is just crazy.

http://en.wikipedia.org/wiki/Chicago_%27L%27

Shortly after its takeover of the 'L', the CTA introduced an express service known as the A/B skip-stop service. Under this service, trains were designated as either "A" or "B" trains, and stations were alternately designated as "A" or "B", with heavily-used stations designated as "AB". "A" trains would only stop at "A" or "AB" stations, and "B" trains would only stop at "B" or "AB" stations. The system was designed to speed up lines by having trains skip stations with fewer passengers while still allowing for frequent service at the heavily-used "AB" stations. The CTA first implemented A/B skip-stop service on the Lake Street Line (now part of the Green Line) in 1948, and the service proved effective as travel times were cut by a third.

[ClintonL94]

Ormeau - Coomera and Helensvale - Robina duplications.  Sod all change to the basic service frequencies, just a bit of reliability.  

Is that right?  Was still half hourly even without the duplications.

Current timetable has a cross between Nerang and Helensvale.  This would have needed to be done at a station with 2 platforms

How did they do the 15 minute peak frequency, or didn't they?

It was half hourly offpeak right through to Robina even with single track beyond Ormeau.  I think the cross was at Helensvale during the offpeak period, with some crosses at Coomera in the peak - I can remember sitting at Helensvale for about 4 minutes on a Robina bound offpeak train.  I think there also may have been some hourly frequency in the middle of the day for a while after the line was first (re)opened in the mid 1990s.

I'm not sure what the peak frequencies were before the duplications, from memory one or two extra peak services were added after the duplication, but not many.

This might help,

Gold Coast timetables..
July 2007
March 2008

[somebody]

What happens when/if the triplication extends to Kingston?

I think that might make david's proposal a lot more doable.

I might have to revise this comment.  The triplication to Kingston will likely only proceed to allow 10 peak minute frequency to/from Varsity Lakes or whatever the current terminus of the Gold Coast line is.  Therefore unless there is some revision to GC service patterns, it would only make sense to also have 10 minute peak frequency from Beenleigh, with express service.  So, if the train from the coast leaves Beenleigh at :07, and the Beenleigh starters leaves at :00 and :10, the stoppers may arrive at Kingston at :13 & :23, with the coast train passing at something like :16.  This means the coast train needs to use the third road and the stopper the middle track from Kingston.  I do not see it being achievable to have much of a counter peak service without quadding in this sort of scenario.  I guess it may be possible to have the train from Beenleigh crossing back to the third road after some station and then running limited stops, but that makes for limited passing opportunities for counter peak trains, especially given the need for the all stopping trains to use the middle track, really.  Seems to be not a happening thing without the quad, and one of some distance too.  Perhaps we don't need the Kingston triple.  I'm starting to hope not.

That makes the question, then, what is the timeframe on a 10 minute Gold Coast peak frequency?  And what is the plan to have "ExpressLink" for stations beyond Loganlea?  Only thing which makes sense to me is forcing the GC trains to serve these stations.

TT keeps talking about the Mandurah line, but did the price tag include resumptions or was that in a reserved corridor.  I do not think we could get a Mandurah line implemented here.

Service speed remains hampered by 19th century alignment from Beenleigh to the CBD & conflicts with the all stations Beenleigh service.

I agree. The stations are also ridiculously closely spaced- about 1km apart, which makes the service unbearably tedious. There is a non-infrastructure solution for this, but it is a political hot potato. If frequencies were increased beyond the 15 minute timetable, skip stop operation would be possible. At the moment, buses are faster in parts- which is just crazy.

http://en.wikipedia.org/wiki/Chicago_%27L%27

Shortly after its takeover of the 'L', the CTA introduced an express service known as the A/B skip-stop service. Under this service, trains were designated as either "A" or "B" trains, and stations were alternately designated as "A" or "B", with heavily-used stations designated as "AB". "A" trains would only stop at "A" or "AB" stations, and "B" trains would only stop at "B" or "AB" stations. The system was designed to speed up lines by having trains skip stations with fewer passengers while still allowing for frequent service at the heavily-used "AB" stations. The CTA first implemented A/B skip-stop service on the Lake Street Line (now part of the Green Line) in 1948, and the service proved effective as travel times were cut by a third.

I don't know how they sped up their service by one third, but I do not see that being achievable here.

[mufreight]

Skip stop operation is a considerable deterrent to service usage, yes it does address in some way the infrastructure deficiencies and by deterring patronage eases the load on rail but then pushes those who have been deterred from using rail onto road causing more congestion and greater cost.
Unfortunately the SEQ bus public transport is not in a position to service the demand that rail skip stop operation would create, not having the road infrastructure needed to allow the operation of the parallel fill in routes that would thus be needed   

[Golliwog]

I'm not a big fan of skip stop. IMO it just makes trying to catch a train more complicated, especially on our network where most lines would also have express services.

Somebody, looking at the Connecting SEQ rail map, the Expresslink trains from Loganlea would from my understanding be stopping all stations outbound from Loganlea, however it also shows these services terminating at Ormeau. They then have the Coastlink services doing basically what the current GC line trains do, with Urbanlink trains inbound from Loganlea as well as running between Coomera and Coolangatta. IMO thats a bit wasteful, it makes more sense to me to run the Coastlink expresses as shown and the Urbanlink trains inbound from Loganlea, but I don't think having the Expresslink trains terminate at Ormeau makes much sense. I would think the network would be better utilised if the Expresslink trains, once they start stopping all stations go all the way to Coolangatta, perhaps replacing the Urbanlink service from Coomera to Coolangatta if the frequency can be high enough. It just makes it easier to travel between stations down that way.

[somebody]

Skip stop operation is a considerable deterrent to service usage, yes it does address in some way the infrastructure deficiencies and by deterring patronage eases the load on rail but then pushes those who have been deterred from using rail onto road causing more congestion and greater cost.
Unfortunately the SEQ bus public transport is not in a position to service the demand that rail skip stop operation would create, not having the road infrastructure needed to allow the operation of the parallel fill in routes that would thus be needed   

I would suggest that it probably depends entirely on the minimum service level.  Notice how the wiki page TT mentioned more or less implies this.  If all stations have a 15 minute service, and the busy ones an 8 minute or so service, then it would be a lot different to busy having a 15 minute service and others a 30 minute service (the current case in Sydney on the Illawarra & Western Lines - and I understand Melbourne has a skip stop in places too).  Although if the infrastructure can support it, you should have the long distance trains missing all the minor stations with additional short working trains IMO.

Somebody, looking at the Connecting SEQ rail map, the Expresslink trains from Loganlea would from my understanding be stopping all stations outbound from Loganlea, however it also shows these services terminating at Ormeau. They then have the Coastlink services doing basically what the current GC line trains do, with Urbanlink trains inbound from Loganlea as well as running between Coomera and Coolangatta. IMO thats a bit wasteful, it makes more sense to me to run the Coastlink expresses as shown and the Urbanlink trains inbound from Loganlea, but I don't think having the Expresslink trains terminate at Ormeau makes much sense. I would think the network would be better utilised if the Expresslink trains, once they start stopping all stations go all the way to Coolangatta, perhaps replacing the Urbanlink service from Coomera to Coolangatta if the frequency can be high enough. It just makes it easier to travel between stations down that way.

So ExpressLink only serves Loganlea-Ormeau and perhaps busy stations inside Loganlea.  Seems pretty weird.

[somebody]

Can we get back to the point here?

If a 10 minute peak Gold Coast service is ever required, this has massive impacts on the Beenleigh line.  I say 9 car trains will head this off, and I guess also removes the requirement to duplicate the bridge over the Coomera River and triplicate to Kingston.

[Golliwog]

But have they mentioned anything yet about upgrading GC stations to be 9 car train compatible? I know they're designing CRR for it, but I'm yet to hear anything about any other stations.

[somebody]

But have they mentioned anything yet about upgrading GC stations to be 9 car train compatible? I know they're designing CRR for it, but I'm yet to hear anything about any other stations.

No I do not think they have said anything about this.  Perhaps they should.

[somebody]

Why can the Gold Coast trains not serve Park Road?

I understood that the platform on the DG couldn't be used for some reason?  Too big a gap or something like that.

Can anyone confirm or deny this one?  I think it is something which something should be done about, even if it is to put in a crossover to access platform 3 from the DG.

I do not see how a reasonable service can be implemented if this limitation still applies.

[somebody]

So are you saying there are conflicting moves even if a new crossover is put in place?

Yes that's correct. The main reason why conflicts occur in the PM peak is because of the GC expresses using the dual-gauge between South Brisbane and Salisbury which is unavoidable.

Below is a diagram of what might occur if PM counter peak was 15 minutes. This use of the tracks would probably occur around after 4:30pm, when 15 minute frequencies of GC begin.



Purple Line is Beenleigh bound, Green Line is Gold Coast bound, Blue Line is Beenleigh to City, Gold Line is Gold Coast to City

Of course, you have the problem of Kuraby terminators as well. But if we pretend that they could magically disappear after terminating at Kuraby (e.g. sending them down to Kingston to turnback) and assuming the current outbond PM peak timetable doesn't change, then a 15 minute frequency counter-peak could work like this:

1) :00 and :30 departures from Beenleigh would remain unaffected
2) :15 and :45 departues from Beenleigh (beginning at 4:15pm and ending at 5:45pm) would have to be held on Platform 2 at Kuraby for approximately 4 minutes extra to allow GC-City services to overtake on Platform 3 (inbound services from GC would pass through Kuraby at about :39 and :09 allowing inbound Beenleigh services to depart at :41 and :11)
3) Departures from Varsity Lakes would be 4 minutes earlier at :48 and :18 beginning at 3:48pm and ending at 5:18pm

Everything would have to run like clockwork as delays on any service could throw everything out. Now that the picture is complete, bring on the 15 minute frequencies!

An alternative here may be to have the Beenleigh trains using the third road, DG and having Kuraby terminators (4tph) assuming responsibility for serving Park Rd-Salisbury + Banoon.  In the PM peak, outbound trains to Beenleigh and VL could run through Kuraby #2, with Kuraby #1 becoming a turnback platform.  Limited stops on the counter peak Beenleigh trains may just make such an option workable.

Similarly in the AM, the stopping patterns ideally should match, and besides, using the crossover at Yeerongpilly is/was a 1 in 8 (is that 15km/h?).  Nothing really stops fast trains from Beenleigh just sticking to the third road/DG, and in fact, it may be possible to have the city bound Kuraby starters using the third road from Kuraby #2.  Timings of this mightn't work out though.

[#Metro]

  Aaargh, network diagrams!!!