Diversion of some AM peak Harbour Bridge buses via the Cahill Expressway

As of Monday 18th February, 2013, selected M2 routes (620, 642, 650, 652) will enter the CBD in the morning via the Cahill Expressway, Transport for NSW has announced.

This is a several years overdue move.  Anecdotal evidence is that this has made an impact on the queues leading in to the York St bottleneck.

The inference I read from this is that these routes will not be removed after the NWRL is implemented.  Given that they would not feed an NWRL station anyway, more like Pennant Hills and Beecroft they are planned to stay and hence a solution needs to be found for them.

So what of the other routes in the North West?  Why would they become feeder routes?

Where would we be if we had the Western Express instead of the Epping Chatswood Rail Link?

Much better off.

I would presume that on opening of a Western Express (WEX a.k.a. CBD enhancement project), the Blue Mountains trains would use it along with the Western Line trains.

Harbour Bridge trains (sector 3):

• 4/hour Epping
• 4/hour Hornsby all to Eastwood, West Ryde, Strathfield, Redfern
• 4/hour Cabramatta via Granville
• 4/hour Glenfield via Regents Park

Moving the South Line to the Harbour Bridge has to be considered a given as it is a faster path on the middle pair of tracks (suburbans) from Strathfield to Macdonaldtown and no conflicting move penalty for doing so as applied in the past with that option.

The above assumes that turnbacks would be constructed at Regents Park (to turn around Bankstown trains) and at Cabramatta, with the Lidcombe turnback becoming unneeded.

A Glenfield via Regents Park to the Harbour Bridge should be somewhat faster than the existing service to Liverpool and help get people out of their cars.  The major limitation with the proposed patterns is that there is little room to allow for growth in patronage on the Western Line, other than with the larger trains proposed, without kicking the Blue Mountains trains out onto the Harbour Bridge, which in turn blocks increases in the Harbour Bridge services.  One solution might be a Homebush-Lidcombe sextuplication, but I have no idea what that would cost.

Outer City Circle, AM peak, would handle all trains from Bankstown, Ashfield, Homebush and Campbelltown via Sydenham.  This would enable simplicity in timetabling and therefore reliability as every train past Erskenville platform 1 would head the same way, however it might be a little controversial as the via Museum path is left with significant unused capacity.  The Erskenville-Sydenham sextuplication would be the next enhancement to increase capacity in this part of the network.

In the PM peak, I would expect that the Campbelltown via Sydenham trains would leave the CBD via Museum, however, the Bankstown trains would be more balanced running via Town Hall at all times.

You can see that the WEX would indeed be well used and the only thing preventing CBD capacity from the west from being fully populated is the lack of rolling stock.  While the via Museum leg of the City Circle remains under utilised, that is left in reserve to allow for increased running via Revesby and Bankstown.  This leg cannot be used for Western Line trains in any event and only 4 trains per hour in the AM peak from the Bankstown line are able to be moved to it and filled in by other trains from the West.

NWRL justification

The justification for the North West Rail Link, contains a couple of interesting points in objective 4 in section 22.3.1:
(a) 29 million trips within 5 years of its opening, which is about half the number of trips currently using the bus system in the relevant region, region 4.  Obviously, people will still continue to use buses into Parramatta and other locations but I expect more than half of bus users are going to or from the CBD-Macquarie Park area.
(b) almost 160 buses are to be removed from the CBD in the morning peak.  This is approximately the amount of buses which enter the CBD in the morning peak from the North West.

I have little doubt that point (b) would not actually happen within the next decade, particularly while the rail-rail interchange at Chatswood remains.

Basically, the justification does not suggest an increase in the public transport usage in the North West.

Northern Beaches BRT study released

Bus Rapid Transit (BRT) has recently been studied for the Northern Beaches.  Long story short, it finds that none of the studied concepts is cost effective.  Why might this be? 

1. 24 hour bus lanes are a bit wasted if off peak services are to be kept mediocre.  I mean, for crying out loud, 20 minute weekend frequency on the L90 (to Palm Beach) dropping back to half hourly evenings and early mornings, with the L88 (to Avalon) not operating is mediocre at best.  In general, 15 minute frequency is what is needed to boost patronage as a 14 minute wait is far more acceptable than a 19 minute one.  Similarly, the L80 (to Collaroy Plateau) reverts to a half hourly 180 on Saturday and the L85 (to Mona Vale via North Narrabeen) is also half hourly.  Those heading beyond Burnt Creek Deviation are best off to avoid the all stopping services as they are about 8 minutes slower from Wynyard to Condamine and King Sts, so effectively there are only 5 buses an hour on this corridor Saturday daytimes with 20 minute service gaps, speaking generally.
2. A 6 lane Spit Bridge is hard to justify as there are already 3 lanes for the peak direction.  This does prevent the limitation of needing to reduce to a single lane counter peak, but it is easy to see that not being justified.  The Roseville Bridge is a reasonable alternative for counter peak travel for half of the Northern Beaches.
3. Option One seems to include the removal of the tidal flow arrangements on the Spit Bridge, with a single lane for cars in both directions.  This is not made clear however.
4. The suggestion of removing the L60 (Chatswood-Mona Vale via Dee Why) is bizarre.

What are alternative proposals? 

1. Have a counter peak bus lane along Spit Rd, preferably extending all the way back to the Warringah Freeway.  This likely would need to be combined with pushing trucks into lane 3 from the start of the Truck and Bus Speed Limit shortly before Medusa St.  There are no right turns until the end of the current speed limit, even if there is one immediately after its end.  This bus lane is important as it allows buses to queue jump the counter peak traffic letting them both serve the counter peak demand better and return for another peak run without requiring more buses to provide the same service.
2. Upgrading the peak direction Transit Lane to a Bus Lane between the Spit Bridge and the Warringah Freeway.
3. Actually establishing some sort of PM peak priority north of Ourimbah Rd.
4. Increase off peak services.  BUZ in Brisbane and Smartbus in Melbourne have both seen high take up of their services.  Metrobus in Sydney has been less successful, even if there are good services in there.  Services stopping at 9pm is no doubt a factor in the lack of success of Metrobus.  This would do much to reduce traffic congestion on the beaches and also help justify 24hr bus lane proposals.
5. More limited stop services outside of peak hour.  It is unclear why the L80 should downgrade to a 180 on weekends, for example.
6. Free transfers.  This will improve the service for all in that people can use a feeder and trunk model to get faster journeys.  Some will prefer a slower single seat, of course.

More generally, it is not entirely clear why median stations would be preferred over kerbside stations, nor is it clear why North Sydney termination with interchange to what are already reasonably full trains would even be proposed.

None of the options in the report break down the costs into their components, largely leaving the reader guessing.

Cityrail timetable proposals in recent news - Part 3

An official video shows that the second phase of this plan involves:

1. A new harbour crossing, connecting to the Metro Pitt alignment.
2. Metro connecting to the Bankstown line and part of the Illawarra line.
3. Removal of Campbelltown via Sydenham. 
4. No increased capacity for the Western, South or North lines beyond what is possible now.
5. Some Western line trains to take the slower path around the city circle.

Almost every aspect of this plan is daft.

1. Sending metro trains down the Illawarra line to Hurstville means that some Railcorp passengers will be taken away from the Eastern Suburbs Railway, to be replaced by no one.
2. Overtaking manoeuvres will no longer be possible.  All trains will likely need to run with the same stopping pattern, presumably Redfern, Sydenham, Wolli Creek, Rockdale, Kogarah, Hurstville and then all stations to Waterfall or Cronulla.  A capacity reducing skip stop is also possible.  It will be exceedingly difficult to operate the South Coast trains with only one pair of tracks the whole distance.
3. Removing the Bankstown line from the city circle removes what is the only swing player between the Town Hall and Museum legs.  This will not be helpful.
4. Removing Campbelltown via Sydenham will result in slower journeys and be less competitive with road transport, particularly with the planned M5 widening.  I do not think that additional frequency through the Airport line adds nearly as much value.
5. Additional Western line paths into the city circle requires them to either add a conflicting move at Macdonaldtown or traverse Strathfield platforms 7 and 8.  The latter will increase journey times, the former would reduce reliability.  There is a good reason why every inquiry has recommended a new CBD path, to prevent such trades from being made.  This new path is being done in such a way as most of the benefit which should be realised is completely wasted.  You might as well run in to the Wynyard platform 1 and 2 dead end as the trams used to.

If a crossing under the harbour is to be done, it really should increase capacity for the other side of Central.  This plan does not achieve that at all.  It wastes some of the presently available capacity.

I hope that we hear nothing further about this plan.

Cityrail timetable proposals in recent news - Part 2

It has recently been announced that the promised plans of double deck services from the North West all the way to the CBD are to be revised.  They are going back to a North West metro, single deck, and only running as far as Chatswood as a first phase.

Firstly, this means that the upper Northern Line will need to revert to running via Strathfield.  Unless it is going to take paths through Town Hall platform 3 from the Western Line it will also mean that the upper Northern Line will terminate at Sydney Terminal.  The only question raised with this is if there is enough capacity between Redfern, Central and Town Hall.  Well in the AM peak 4/6 Bankstown line trains add to this capacity.  This should be relatively simple to expand to the other two with a timetable review.  This capacity should be relatively easy to provide in the AM peak, and the PM peak is far less congested.  Reversing the direction of the Bankstown line around the city circle full time may well be a reasonable option, as might Melbourne weekend style Bankstown-Town Hall-Museum-Bankstown full time.

Secondly, it means no expansion to Railcorp and in fact a slight contraction as they will be removed from the Epping-Chatswood Rail Link.  In my opinion, this is real positive to the proposal.  I'm really not sure how long we can go having Railcorp consume $3 billion per year and rapidly rising.

Thirdly it means an interchange at Chatswood for trains which are allegedly already full.  However, they aren't as full as trains on other lines and the upper Northern Line passengers are to be removed from this patronage.  It will also be possible to increase the numbers of trains on the North Shore by up to 6 per hour.  There will be plenty of capacity for passengers from the North West who will have a cross platform interchange at Chatswood, much like Glen Waverley line peak hour passengers in Melbourne who need to interchange at Richmond to reach city loop stations.  Hardly the end of the world.

All in all the first phase of these changes is quite good, which is something I will not say about the second phase in my next blog post.

Cityrail timetable proposals in recent news - Part 1

A Sydney Morning Herald article has reported that it is being suggested for a number of stations to lose their access to Town Hall, but the Cumberland Line is to be given an upgraded service and run through to Richmond.  These changes are sure to be unpopular and reduce the effectiveness of Cityrail, so the question is why would it even be considered?  One word: sectorisation.  This means that parts of Cityrail can function independently of the whole, as has been all but achieved with the Illawarra and Eastern Suburbs lines.  Cityrail branded this concept clearways a number of years back.

Basically, to have a strict sectorisation the Cumberland line would need to either be completely removed, or pushed into what is known as sector 2, meaning lines accessing the City Circle.  This means that lines through Parramatta platforms 3 and 4 need to go to sector 2, and given that it is not possible to come from the Richmond line without accessing these lines, the Richmond line needs to go along for that ride.  Since trains serving Harris Park, Wentworthville, Toongabbie and Pendle Hill normally only use Parramatta platforms 3 and 4, these trains would no longer be able to reach Central.  Similarly, Clyde, Auburn, Lidcombe and Burwood would no longer have a single seat service to Parramatta.  Clyde stops would need to be added to selected south line trains.  The most popular peak pattern - All to Redfern/Parramatta/Blacktown/Mt Druitt all to Penrith would need to be slowed down significantly with probably 6 more stops.  It's not entirely clear why the lower Northern line cannot run through Town Hall with sectorisation however, even with the conflicting move* between outbound interurbans and inbound suburbans north of Strathfield.  It is also unclear how Blue Mountains line trains would operate at all if this is done.

This is sectorisation becoming the enemy of Cityrail doing a good job, rather than its friend, which it usually is. I feel this is being leaked now to soften up users of Cityrail for a timetable which isn't going to be liked.  It is certainly true that the Western Line needs its own path through the CBD though, which should have been built before the Epping Chatswood Rail Link was.

What alternatives are there for 2013?
Alternative One:
Increase running into Sydney Terminal for the Western Line.  It is unclear why there are still 3 (both directions combined) Springwood trains per day crossing the Harbour Bridge.  This was done before the ECRL freed up a few paths across the Harbour Bridge so there is precedent.

Altenative Two:
Run additional lower Northern Line trains into Sydney Terminal. This is a satisfactory option for increasing the Northern Line which is incredibly overcrowded but does nothing for the Western Line, which is nearly as bad.

Alternative Three:
Run all lower northern line trains into Sydney Terminal. The existing trains should not be damaged in this way, even if increased in frequency.  Varied stopping patterns cannot reasonably apply for the Western Line between Redfern and Parramatta so all my comments above would apply in that zone and about the Blue Mountains Line trains being virtually unable to run.

Alternative Four:
Connect the lower Northern line the City Circle, moving some or all of the inbound AM Bankstown line trains to the Museum.  This has a conflicting move* at Macdonaldtown and thus would surely reduce reliability to pre-2005 levels so has a snowballs' chance of proceeding.

* A conflicting move is where one train needs to get in the way of a train on another track, blocking its progress.  While it can be timetabled, things don't always run to timetabling.  It ultimately reduces reliability.

UPDATE 22/6/2012: Or here's a better idea: Terminate the Cumberland line at Parramatta from the south.  This means that Parramatta platforms 3 and 4 can become part of sector 2 and Harris Park commuters need to backtrack to Parramatta to reach the CBD.  This is not really a big problem as they are likely to have a faster trip to the city by doing this in any event.  This means that all tracks and platforms at Westmead and west are available to sector 3 trains without violating sectorisation and it also means Cumberland trains don't run relatively uselessly all the way to Blacktown.

Second rail harbour crossing back on the agenda

The Sydney Morning Herald has again reported that a new harbour crossing is back on the agenda.  The SMH reports that the advice to Gladys Berejiklian is this is needed for natural growth and the NWRL.  Well, sort of.  It is only needed in this or the next decade because of the NWRL.

It was always disingenuous and/or dishonest to suggest that the NWRL without a second harbour crossing was a reasonable policy, but it was a winning strategy at the 2011 NSW election.  In my opinion Infrastructure Australia was absolutely correct to rebuff the current NWRL plan.

Some people have suggested that signalling improvements could delay the need for this.  Well that may be, but it would not come cheap and nor would it remove some remaining conflicting moves particularly at Macdonaldtown.  Signalling improvements also would not help to remove the south line from the need to pass the platform faces between Summer Hill and Macdonaldtown - these lines are slower than the centre pair of tracks, which are slower than the northern pair of tracks.  There needs to be a study on the bang for buck of this particular option.  Personally, I believe it will defer the need for track amplifications only slightly so should not proceed.

What about single deck?  Well while you might have more trains, the number of seats on the line would be reduced for sure.  It is unclear whether or not the number of standing spaces would be increased by enough to compensate, but I think it is fair to say that reduced seating is against what Sydneysiders want.  There are similar issues with increasing the number of doors per carriage.

The SMH also reports that the "City Relief Line" is dead.  Well, that's creative politicking right there.  Transport for NSW report that all options feature a "CBD enhancement" which is basically the city relief line by another name.  As for the suggestion the under the harbour line might connect to the Illawarra, that is insane, and a waste of web bandwidth to debunk.

De-bottlenecking York St

Yesterday I blogged that the NWRL cannot be relied on to fix this problem.  So what other fixes are possible?

1. The southbound M40 would benefit from using the Cahill Expressway to Elizabeth St.  This will avoid the Park St interchange, but who cares?
2. The Pacific Highway is over serviced and IMO the best solution is chopping off the northern part of the M20
3. STA routes E86-9 have been removed from York St to use Grosvenor and George Sts in the AM peak.  This treatment should be spread to all AM peak only Hillsbus routes - 613, 614, 616, 617.
4. A double decker bus meeting all ADRs (Australian Design Rules) has come on to the market.  Yet this has not been ordered by any NSW operator to the knowledge of the author.
5. Carrington St is not used as a drop off for Wynyard terminating buses.  This is presumably due to the difficulty of accessing Clarence St via Margaret St.  The fix for this problem would be to reverse an eastbound lane of Margaret St between Clarence and York Sts.  This would allow these buses to return to their depots more quickly or perhaps in time to make another peak run.

Put simply, there are a number of relatively simple fixes, just not actioned.

Harbour Bridge bus congestion - will the NWRL fix it?

Recently, 358 buses were timetabled to arrive at Wynyard from the Harbour Bridge in the hour from 8am to 8:59am on a work day.  79 (or 22%) of these buses are M2 services.  So one might expect a significant reduction in congestion once you consider queueing theory's prediction that a small change in supply or demand can have a dramatic effect on waiting times.  This prediction has a number of problems:

1. The remaining services can be expected to continue to grow.  Growth of 2.5% p.a. over 10 years would put us basically back where we started
2. Cityrail's network could not handle this many additional passengers travelling over the Harbour Bridge without a lot of pain.  Transport for NSW's preferred plan is to run the trains for the upper Northern Line into Sydney Terminal, however two trains per hour via Gordon must also be trimmed.  These changes seem extremely unlikely to survive political interference, and the alternate plan of terminating every second NWRL train north of Sydney Harbour seems likely to proceed
3. If only half of the promised 8 per hour trains actually reach Wynyard, it is likely that the short workings will not add much value.  Their value would be increased by the previously proposed quad between St Leonards and Chatswood.  This would allow people bound for stations up to St Leonards to use the short workings, and running the full distance trains express between Epping and St Leonards would also channel people who can use the short workings on to them.
4. Integrated fares have not been achieved in Sydney, and even if they were it is unclear that people from the Hills District would be willing to use a feeder bus to reach the station.  If they will not, then that means very expensive parking must be provided, a reduction in the market penetration of public transport or the buses along the M2 to the city and North Sydney would need to be retained in a reasonably large number.
5. An additional 25 or so buses reach North Sydney from the M2 between 8am and 8:59am on a working day.  It is estimated that between the Wynyard direct and North Sydney/Milsons Point buses that 4 trains per hour reaching North Sydney and Wynyard would be insufficient to meet demand.

So it can be seen that the North West Rail Link can not be relied on to fix bus congestion on the Harbour Bridge.  A subsequent post will go in to solutions to fix these problems.