POLL: Beenleigh line upgrade

[somebody]

For more info:
David's plan: http://railbotforum.org/mbs/index.php?topic=4506.0 Basically, not maintainable counter peak and not pretty for reliability issues
Somebody's plan: http://railbotforum.org/mbs/index.php?topic=5569.0 Still not pretty for reliability but maintanable counter peak.  Requires changing the Gold Coast trains back to serving Yeerongpilly.  Moorooka could also be a viable option.
Option 4 also in the thread for david's plan but requires infrastructure for reliability.

[#Metro]

My head is spinning... how about a timetable mock up for the off peak between 12 noon and 1pm on a weekday.

[somebody]

My head is spinning... how about a timetable mock up for the off peak between 12 noon and 1pm on a weekday.

Which option?

[david]

Done a quick mock-up of my proposal

Gold Coast/Beenleigh to City




City to Gold Coast/Beenleigh



...and the line diagrams

Salisbury to Sunnybank


Sunnybank to Kuraby


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

Note that the Gold Coast-bound train crosses from Up Main to Down Main around :13/:43
The inbound Beenleigh train crosses from Third Road to Down Main at :20/:50
Hence, there is no conflicting move.

The only flaw in my plan (as can be seen on the City to Beenleigh/GC timetable) is that the GC trains will be slowed down by the Beenleigh train in front until it crosses onto the Down Main. On my schedule, it has both the Beenleigh and GC trains arriving at Coopers Plains at :15/:45. This is impractical in the real-world.
Of course, one could easily add a couple of minutes onto the GC run.

At the Beenleigh end of things, I propose the following:

- Beenleigh train arrives at :19/:49 on Platform 1 and remains there (or can move into siding) until departure at :30/:00

- Beenleigh train arrives at :34/:04 on Platform 1 -> moves into siding -> Gold Coast train arrives Platform 1 at :40/:10 -> Train moves from siding back to Platform 1 to form :45/:15

This would give an 11 minute dwell. It also explains why I can't budge on the timings of the City to Beenleigh services and would prefer to add a few minutes to the GC run when it comes to the Coopers Plains problem I explained above.

Hope this clears up any confusion that surrounded my proposal.

[Gazza]

Quick question, but why is Park Rd no longer served by GC trains?

[david]

Under my proposal, inbound GC trains would have to use the dual gauge line full-time, hence why it wouldn't be able to stop at Park Road.

[somebody]

Under my proposal, inbound GC trains would have to use the dual gauge line full-time, hence why it wouldn't be able to stop at Park Road.

To add to this, platform 4 at Park Rd is locked off, and AIUI deemed unusable.  However, platform 3 is in regular use.  You could add a crossover from the DG to platform 3 and then have the trains serve Park Rd.  They could also serve Park Rd in peak with this.

[Gazza]

Ah gotcha....Kind of bugs me they spent hundreds of millions on a busway connection from the station to the Uni, yet many passengers still cant take advantage of this because the platforms aren't right.

How much could a crossover cost anyway?

[Stillwater]

We should prepare, and send to QR, a list of crossovers and other minor track enhancements required to tweak the network in order to achieve greater frequency -- starting with the Beenleigh Line.  Some targeted expenditure on relatively small measures with big benefits could be highly cost-effective.

[#Metro]

David's plan has more info... and look at Beenleigh station!
You would be able to feed buses into there with no problem!


We must use what we have much better.

[somebody]

Ah gotcha....Kind of bugs me they spent hundreds of millions on a busway connection from the station to the Uni, yet many passengers still cant take advantage of this because the platforms aren't right.

How much could a crossover cost anyway?

But are there really that many students commuting from the Gold Coast?

My guess is that a crossover would be around the $1m mark.  Signaling is part of the expense.

[#Metro]

What do those small numbers in the network diagrams like 25/25 or 16 mean?

[colinw]

That is a great proposal that has been worked up by David, and certainly gets my vote.

The only flaw in my plan (as can be seen on the City to Beenleigh/GC timetable) is that the GC trains will be slowed down by the Beenleigh train in front until it crosses onto the Down Main. On my schedule, it has both the Beenleigh and GC trains arriving at Coopers Plains at :15/:45. This is impractical in the real-world.
Of course, one could easily add a couple of minutes onto the GC run.

Ah, that explains why quad to Banoon has been identified in previous studies as being necessary for full 15 minute or better frequencies.

How far before Coopers plains does your timetable have GC trains running on the yellows of the preceding all stations Beenleigh?  I'm wondering if there is potential to fix this problem with a crossover back toward Moorooka or Rocklea combined with bidirectional signalling (if it doesn't already exist) to allow the GC train to overtake using the down line. (Or does that introduce a conflict with a Beenleigh to City service?).

How did you model this?  Spreadsheet, or something more sophisticated.  Or did you actually do the diagrams?

[colinw]

What do those small numbers in the network diagrams like 25/25 or 16 mean?

It is a measure of the angle of the turnout at a crossover or junction, and can be used to read whether turnouts are low or high speed.  Higher numbers are faster.

12 = 1 in 12 turnout (= 25 km/h)
16 = 1 in 16 turnout (= 50 km/h)
25 = 1 in 25 turnout (= 80 km/h).

Thus at the locations around Salisbury where David's plan has services using low speed crossovers marked '12', the trains would be slowing to 25 km/h to use these crossovers.

Upgrading to a higher speed crossover could make things more efficient, assuming there is room to fit in a higher speed crossover, as they are longer due to the less sharp divergence.  There are significant costs involved in such upgrades as the overhead has to be adjusted as well, possibly involving fitting new masts or portals.  Not cheap!

[somebody]

I'm wondering if there is potential to fix this problem with a crossover back toward Moorooka or Rocklea combined with bidirectional signalling (if it doesn't already exist) to allow the GC train to overtake using the down line. (Or does that introduce a conflict with a Beenleigh to City service?).

Yes. - that introduces a conflict for Beenleigh-City, or at best reduces the reliability of an already difficult to maintain timetable as a track is used by trains running every 15 minutes in both directions for a longer period.

Skipping the 7 stations Dutton Park-Salisbury is about the minimum for a train from the Gold Coast which is 3 minutes behind the train from Beenleigh to get ahead before Park Rd/South Bank.  There are cross overs at Yeerongpilly, but I do not think they could be used.

12 = 1 in 12 turnout (= 25 km/h)
16 = 1 in 16 turnout (= 50 km/h)
25 = 1 in 25 turnout (= 80 km/h).

Great info there!

Thus at the locations around Salisbury where David's plan has services using low speed crossovers marked '12', the trains would be slowing to 25 km/h to use these crossovers.

Upgrading to a higher speed crossover could make things more efficient, assuming there is room to fit in a higher speed crossover, as they are longer due to the less sharp divergence.  There are significant costs involved in such upgrades as the overhead has to be adjusted as well, possibly involving fitting new masts or portals.  Not cheap!

That's not the main problem.  The main problem is that there is a conflicting move on the middle track between Coopers Plains and Salisbury.  Although perhaps these occur at different times on the timetable, by enough that it doesn't cause a huge problem.

EDIT: clarity

[david]

The only flaw in my plan (as can be seen on the City to Beenleigh/GC timetable) is that the GC trains will be slowed down by the Beenleigh train in front until it crosses onto the Down Main. On my schedule, it has both the Beenleigh and GC trains arriving at Coopers Plains at :15/:45. This is impractical in the real-world.
Of course, one could easily add a couple of minutes onto the GC run.

Ah, that explains why quad to Banoon has been identified in previous studies as being necessary for full 15 minute or better frequencies.

Yes. I'm a little worried that CRR is being touted as the solution for 15 minute frequencies on the Beenleigh/Gold Coast line. The is simply not the case from my modelling. Personally, I'd like to see a quad to Kuraby and triplication to Beenleigh, but that's probably never going to happen. IMO, CRR will only provide more flexibility for the scheduling of Cleveland services.

How far before Coopers plains does your timetable have GC trains running on the yellows of the preceding all stations Beenleigh?  I'm wondering if there is potential to fix this problem with a crossover back toward Moorooka or Rocklea combined with bidirectional signalling (if it doesn't already exist) to allow the GC train to overtake using the down line. (Or does that introduce a conflict with a Beenleigh to City service?).

I'm not quite sure what you mean by this, but I'll assume that you mean yellow signals. In that case, you're correct in that the GC train will start to hit single yellow signals between Moorooka and Rocklea. This occurs around :11/:41. There would be a possible conflict with the inbound Beenleigh service, since there's only really a 3 minute leeway in my timetable (Beenleigh to City train departs Rocklea at :08/:38)

How did you model this?  Spreadsheet, or something more sophisticated.  Or did you actually do the diagrams?

Yep, it was all Excel. It's amazing what you can do with some clever algorithms.

[#Metro]

Yes. I'm a little worried that CRR is being touted as the solution for 15 minute frequencies on the Beenleigh/Gold Coast line. The is simply not the case from my modelling. Personally, I'd like to see a quad to Kuraby and triplication to Beenleigh, but that's probably never going to happen. IMO, CRR will only provide more flexibility for the scheduling of Cleveland services.

It's called a SMOKESCREEN.
Find a convenient alibi and use it to starve off real change.

15 minute frequency is possible on much of the network with minimal or no new infrastructure and no new trains.

[Golliwog]

I thought CRR was actually going to do some quading of the GC/Beenleigh line? Although I think it was only as far as Salisbury IIRC from the map on the CRR website.

[somebody]

Yes. I'm a little worried that CRR is being touted as the solution for 15 minute frequencies on the Beenleigh/Gold Coast line. The is simply not the case from my modelling. Personally, I'd like to see a quad to Kuraby and triplication to Beenleigh, but that's probably never going to happen.

CRR would allow 15 minute frequency to Kuraby with no additional crossovers or impact on the Gold Coast line.  But that's the only thing they could be thinking of.

[colinw]

Yes. I'm a little worried that CRR is being touted as the solution for 15 minute frequencies on the Beenleigh/Gold Coast line. The is simply not the case from my modelling. Personally, I'd like to see a quad to Kuraby and triplication to Beenleigh, but that's probably never going to happen. IMO, CRR will only provide more flexibility for the scheduling of Cleveland services.

I was wondering the same thing, which is why I brought it up.  CRR will certainly alleviate the Merivale Bridge / Roma St choke point but it seems to me that there are also capacity constraints between Yeerongpilly and Coopers Plains, and again beyond Kuraby.  Also, not helped by the lack of crossovers between Coopers Plains and Kuraby and the layout at Kuraby which causes conflicts for terminating trains.

I'm not quite sure what you mean by this, but I'll assume that you mean yellow signals. In that case, you're correct in that the GC train will start to hit single yellow signals between Moorooka and Rocklea. This occurs around :11/:41. There would be a possible conflict with the inbound Beenleigh service, since there's only really a 3 minute leeway in my timetable (Beenleigh to City train departs Rocklea at :08/:38)

Yes, that was what I meant.  Seems like quad from Yeerongpilly to Salisbury (as proposed as part of CRR) is necessary then, probably to Coopers or Banoon as previously modelled in a study.

[somebody]

The problem with the plan outlined by david is that it has no operating margin.  When the southbound Beenleigh train is delayed by a few minutes by Salisbury, the southbound coastie is also delayed by a few minutes.  This also means that the northbound Beenleigh train cannot access Salisbury without further delaying the two southbound trains.  If the decision to prioritise the northbound services is not taken, the northbound coastie will also be delayed.  This isn't a nice decision.  Basically delay 4 trains by a few minutes or 2 trains by quite a bit.  Also, once the northbound coastie is delayed, that will probably delay the next southbound one at the Airport with the single track.  Likely, there will also be delays for the next Beenleigh service that the Beenleigh train does as the turnback arrangements there do not allow for fat turnbacks with a 15 minute frequency.

There are similar, although less severe issues for the southbound trains at Kuraby and the northbound trains at Coopers Plains.

I think it will be a cold day in hell when QR implements this plan on current infrastructure.  We can't even get a 15 minute frequency to Petrie!

The Kingston triple does add some operating margin, so long as a couple of extra crossovers are put in at around Sunnybank and probably another one at Park Rd.  I strongly feel that this is the sort of direction we need to be directing our efforts, if we feel that a 15 minute frequency all the way to Beenleigh is a priority.

[david]

The problem with the plan outlined by david is that it has no operating margin.  When the southbound Beenleigh train is delayed by a few minutes by Salisbury, the southbound coastie is also delayed by a few minutes.  This also means that the northbound Beenleigh train cannot access Salisbury without further delaying the two southbound trains.  If the decision to prioritise the northbound services is not taken, the northbound coastie will also be delayed.  This isn't a nice decision.  Basically delay 4 trains by a few minutes or 2 trains by quite a bit.  Also, once the northbound coastie is delayed, that will probably delay the next southbound one at the Airport with the single track.  Likely, there will also be delays for the next Beenleigh service that the Beenleigh train does as the turnback arrangements there do not allow for fat turnbacks with a 15 minute frequency.

I did try and factor in approximately 5 minutes for late running services:

Note that the Gold Coast-bound train crosses from Up Main to Down Main around :13/:43
The inbound Beenleigh train crosses from Third Road to Down Main at :20/:50

Is a 5 minute operating margin too little?

I think it will be a cold day in hell when QR implements this plan on current infrastructure.  We can't even get a 15 minute frequency to Petrie!

Now now...let's not get all pessimistic here. Perhaps someone looking at this is from Queensland Rail, and take a look at this and go "we've never even considered this!" and tweak it a little to suit their operational needs. That's the only reason why I'm putting it out there.

The Kingston triple does add some operating margin, so long as a couple of extra crossovers are put in at around Sunnybank and probably another one at Park Rd.  I strongly feel that this is the sort of direction we need to be directing our efforts, if we feel that a 15 minute frequency all the way to Beenleigh is a priority.

From my modelling, the main bottleneck at the moment preventing 15 minute frequencies is between Salisbury and Banoon. Every combination I've tried results in some sort of clash around Coopers Plains. I'm currently working on a model whereby crossovers are added south of Coopers Plains, which will give a greater operating margin and reduce conflicts between northbound and southbound traffic.

[david]

Assuming that there was some spare cash lying around to pay for 2 new crossovers, I've remodelled my plan for 15 minute frequencies to Beenleigh to increase the operating margin and hence increase reliability.

City to Beenleigh/GC




Beenleigh/GC to City - this is unchanged



and the line diagram:


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

Changes made:
- Two new crossovers added south of Coopers Plains. Note that these crossovers will not conflict with one another (like the crossover between Corinda and Sherwood)
- Inbound Beenleigh-City trains now cross from Third Road to Down Main before Coopers Plains
- Gold Coast train crosses to from Up Main to Down Main after Coopers Plains

I have managed to remove conflicts between inbound and outbound services that existed in my original plan.

However, in the event that either the Beenleigh-City or City-Beenleigh train is delayed, this will result in the Airport or Gold Coast train (respectively) being delayed. With that in mind, I have made allowances at Helensvale (3 minutes) and at the International Airport (2 minutes), plus all of the fat I've added between South Bank and Bowen Hills.

[#Metro]

This is great. 15 minute clockface timetable as well.

Just a question- you have all the GC trains run express past Park road in the off peak. Does your idea work if they all stop there?

[somebody]

I did try and factor in approximately 5 minutes for late running services:

Err, no.

The southbound services have the Coasties 2 minutes behind the Beenleigh trains at Coopers Plains.  Normal margins are 3 minutes.

You don't suggest a time at Kuraby, but I would suggest that this is also of the order of 3 minutes southbound.

Northbound, margin is also only 2 minutes at Coopers Plains.

[david]

Just a question- you have all the GC trains run express past Park road in the off peak. Does your idea work if they all stop there?

Yes they can but, unfortunately, there needs to be a crossover installed between Dutton Park and Park Road for inbound GC trains to stop at Park Road. Either that or fix up Platform 4.

Err, no.

The southbound services have the Coasties 2 minutes behind the Beenleigh trains at Coopers Plains.  Normal margins are 3 minutes.

You don't suggest a time at Kuraby, but I would suggest that this is also of the order of 3 minutes southbound.

Northbound, margin is also only 2 minutes at Coopers Plains.

I'm assuming that you're referring to my new proposal, even though my comment was aimed at my old proposal. With the new proposal, you're right that at Kuraby, there is a 3 minute operating margin. Unfortunately, I cannot budge with the margins at Coopers Plains. If I increase the operating margin at Coopers Plains, this decreases the margin at Kuraby. I think that its more important to have the greater operating margin at Kuraby than at Coopers Plains.

Inbound, I have set up Coopers Plains as like Central Platform 3/4, and they run it at an operating margin of 2 minutes all the time, so I'm assuming it's fine to do this at Coopers Plains as well.

[#Metro]

Yes, I have often wondered it it was possible to fix platform 4.

[somebody]

Your original proposal had trains arriving at Coopers Plains at exactly the same time.  That was clearly impossible without doing something ridiculous.  Obviously, it had to be re-timed.

3 minutes would generally be regarded as the minimum headway for a timetable.  This proposal is probably only doable with some spots of 2 minutes.  Not trying to put it down, but why should we propose something which is very difficult to do when we cannot get them to do things which can easily be achieved?  Especially when it has already been proposed.

[david]

I suppose what I was trying to say is that it is impossible to have 3 minute headways if you want 15 minute frequencies on the Beenleigh line with the current infrastructure. We would need at the bare minimum either quadding between Park Road and Sunnybank or triple to Kingston.

Some may be wondering whether 15 minutes to Kuraby would be better. From my modelling, we would encounter the same problems with tight headways, not to mention the conflicting move required at Kuraby for it to turnback. It's easier to go all the way to Beenleigh.

[#Metro]

Thanks for your contributions David. It's easy to see things when you do up the timetable and the network diagrams.. 

[#Metro]

Mufreight, what is your opinion on these proposals? Do they have sufficient operating margin?

[somebody]

I suppose what I was trying to say is that it is impossible to have 3 minute headways if you want 15 minute frequencies on the Beenleigh line with the current infrastructure. We would need at the bare minimum either quadding between Park Road and Sunnybank or triple to Kingston.

Exactly.

Some may be wondering whether 15 minutes to Kuraby would be better. From my modelling, we would encounter the same problems with tight headways, not to mention the conflicting move required at Kuraby for it to turnback. It's easier to go all the way to Beenleigh.

On currrent infrastructure, that is true in one direction, and it's probably even worse.  Hence the proposal of a crossover to allow the the southbound Kuraby train to get out of the way of the following Coastie somewhere like Altandi.  It would make the 2tph Kuraby trains needing to use an affective single track Altandi-Kuraby-Altandi, but this seems to be doable to me.  Main problem with that proposal is that it isn't doable now.

[O_128]

would running gold coast trains via tennyson fix anything?

[somebody]

would running gold coast trains via tennyson fix anything?

The government's re-election chances?

It would help with the conflicts though.

[#Metro]

Is it shorter or longer via Tennyson?

[colinw]

Anyone got a copy of the timetable from a few years ago on hand?  There was one timetable that had a small number of Gold Coast via Tennyson services.

[somebody]

Is it shorter or longer via Tennyson?

Longer.  Hence electoral suicide.

[mufreight]

Unfortunately the Tennyson line can not be acessed from the dual gauge line.
Had the planned provision of a standard gauge connection into the Mooloobin yards for market traffic been built prior to the construction of the Fairfield Road overpass things would be different but once again thanks to the beancounters not only was the freight traffic at that time lost but also now the potential to operate northbound passenger and freight services from the dual gauge line through Tennyson.

[somebody]

Unfortunately the Tennyson line can not be acessed from the dual gauge line.
Had the planned provision of a standard gauge connection into the Mooloobin yards for market traffic been built prior to the construction of the Fairfield Road overpass things would be different but once again thanks to the beancounters not only was the freight traffic at that time lost but also now the potential to operate northbound passenger and freight services from the dual gauge line through Tennyson.

Huh?  For NG, there are points from the DG and from the down track.  There is also a crossover to the up track.  What is lacking is two roads between Moorooka and Tennyson.  Trains from the DG also do not need to conflict with the trains on the down to Yeerongpilly.

The current timetable even has a Corinda-Moorooka service at 7:27am.  I understand that game specials have operated between Robina and Milton via Tennyson in the past also.

[colinw]

Unfortunately the Tennyson line can not be acessed from the dual gauge line.

Yes it can.  From the south - here.
From the north - here.
http://www.queenslandrail.com.au/NetworkServices/Documents/Cleveland%20Branch%20and%20Beenleigh%20Line%20-%20NAG%20017-16.pdf

Between Yeerongpilly & Acacia Ridge was dual gauged first, in the early '90s, as a non-electrified freight connection between the Tennyson Line / Moolabin and Acacia Ridge.  Electrification and dual gauge Yeerongpilly to South Brisbane happened later, in time for the opening of the Gold Coast line first stage to Helensvale.