Thursday, December 23, 2021

UN 117: The bus and tram network the Metro Tunnel needs (Part 3 - central area)


Parts 1 and 2 discussed revised bus networks for the stations feeding the Metro Tunnel from the west and east respectively. 

The concept of revamped bus and tram networks to coincide with the Metro Tunnel's opening was raised in Infrastructure Victoria's 30 year strategy. This received state government endorsement through the Victorian Infrastructure Plan. The key now is to ensure that it happens with a significant uplift in service. And given the magnitude of doing the job properly planning really needs to start now.  

What about on-street transit for the central Metro Tunnel area itself?

We're talking about five new stations here. These include State Library near Melbourne Central and Town Hall near Flinders Street. And three others (Arden, Parkville and Anzac) at currently non rail -served precincts. 

These should trigger an overhaul of bus and (especially) tram networks. Not so much because of a need to access the stations but because the Metro Tunnel will relieve stress on Melbourne's busiest tram corridor and enable redesigned routes with flow-on benefits across the network. 

Already has state government support

The good news is that, like buses, tram network redesign also has Victorian Infrastructure Plan endorsement. Indeed Metro Tunnel works is delivering a Park St link and budget funds have been committed to planning tram reform. So it looks like a goer. It should even be possible to do much with no or only a small increase in tram service hours if network overlaps are reduced and/or speeds are increased. 


Department of Transport too

Enabling tram network reform with better efficiency and distribution of service is a key rationale for the Metro Tunnel (2016 business case here).  Furthermore, the Department of Transport (then as Transport for Victoria) started planning this well before the IV report and the government's Infrastructure Plan response. You can see an indication of their thinking in the Why can't we just?  presentation from 2017.  

Its 'layered planning model' (Slide 6) describes tram route, frequency and span adjustments as 'easy'. Cost is stated as being low, being in the 'hundreds of thousands' or 'millions'. Slide 12 below mentions the St Kilda Rd corridor and the potential for trams from the south-east to be rerouted via Park St to serve faster growing destinations in the CBD's west. Such rerouting is assisted by the Metro Tunnel's 'stress relieving' role mentioned earlier. 


The availability of the above is noteworthy as the Department rarely publicly engages on network planning matters until work is well advanced. While there isn't always a lot of detail on specific measures, Infrastructure Victoria, with its legislated independence, tends to be more forward on these matters. 

Taking a step back

Let's put individual projects and modes to one side to consider broader CBD transit network issues. A visitor to the Hoddle Grid would not be mistaken for thinking it's paradise with a dense grid of frequent (and free!) trams. And they wouldn't need to walk far before finding a station on its periphery due to the City Loop. 

But you don't need to go far outside it for the story to change. A snapshot of key issues is below:

Strengths
- Focal point for train and tram infrastructure with access from much of state
- Quite good connectivity between trains and trams
- Good integration of trams in the urban fabric of Hoddle grid
- Frequent buses from some important directions

Weaknesses
Complexity of City Loop (eg 2/4 portals still reverse at midday) makes it unsuitable for short trips 
- Complexity of trams in CBD with service maldistributed relative to activity
- Some tram network 'missing links' including to dense areas (eg parts of Spencer St)
- CBD bus corridors are complex, invisible and integrate poorly with trains and trams.  
- Non-radial trips just outside the Hoddle grid often difficult due to inconvenient network design
- Frequencies less than turn-up-and-go on individual routes (eg trains 20 min, trams 12 min) 
- Free Tram Zone unnecessarily crowds and slows trams while leading to underuse of trains and buses

Opportunities
- Metro Tunnel will enable improved trains including a consistent City Loop pattern on all lines
- Metro Tunnel creates opportunities to redistribute tram resources & rethink the network
- A flattened peak and post COVID recovery also creates opportunities to rethink the network

Threats/risks
- A post COVID slump with a reduced central area activity may make reform seem less urgent
- Metro Tunnel implemented without supporting network reform 
- Current interest in network reform dissipates with little achieved
- Good reform might get done but is stupidly reversed (eg recent rerouting of Tram 12 )

6 big moves to transform inner area transit

Now we know the weaknesses, let's focus on the opportunities. The following six identified are either made easier by the Metro Tunnel or support usage of its stations. None are original but I think they'd make a difference and maximise Metro Tunnel benefits.  

1. Go West: Shift some trams from Swanston St to the western part of the CBD

The graph below, from the Metro Tunnel Business Case, tells the story. William St has lots of jobs nearby but few trams. Spencer St also receives relatively low service. Trams from the south-east are  overwhelmingly concentrated on the Swanston St spine. The Metro Tunnel will provided the needed capacity to enable a westward shift so that more trams can run via William and Spencer Street. 


St Kilda Rd currently has eight tram routes (1, 3/3a, 5, 6, 16, 64, 67, 72) from the south that operate through to Swanston St. At least one of these should be rerouted via William St with another via Spencer St to better distribute services. The latter is why the proposed Park St link is important. 

A tram network reconfiguration has several potential flow-on benefits. For instance new train/tram connectivity at Anzac Station. Fewer routes in Swanston St capacity exists for those that remain to be lifted from their current 12 minutes interpeak to a genuine turn-up-and-go frequency.

The ability to do this is further enhanced if there are more through routes with fewer overlaps. For example William Street's extra route from the south-east could form a 57 rerouted from Elizabeth St with this and the 58 providing a William Street super-frequent pair from Park St to Queen Victoria Market. As with all these things cascading effects need to be considered. For example the need to retain frequency on Elizabeth St, such as by providing short trips in a route's busiest portion. 

2.  Connecting the neglected north-west

The north-west of the CBD, around Spencer and La Trobe St has had huge recent development. However its trams haven't kept pace. And poor accessibility to the New Quay area has seen businesses struggle. Page 57 of the Metro Tunnel Business Case maps this accessibility deficit starkly. 

Key issues include (i) The absence of a simple continuous tram route up Spencer St, unlike other CBD streets, (ii) No full-time tram connection from New Quay to La Trobe St destinations including Flagstaff, Melbourne Central and the Tunnel's State Library Station and (iii) No easy access to New Quay from Footscray.


Potential solutions include: 

- Permanently reroute Tram 12 via Spencer St and LaTrobe St as was temporarily done in 2020 during COVID and unaccountably reversed. This change had many network benefits including enhanced connectivity to regional trains and Skybus, more services on Lonsdale St, simpler services on Spencer St, etc.



- Provide a simple west-east connection from Waterfront City via LaTrobe Street operating full time.  This would supplement the existing limited service City Circle (Route 35). There's at least two potential approaches, one radical and one not. The radical approach involves rerouting the 86 tram to shorten its route kilometres and remove three turns. Route 30 would be deleted entirely with resources saved used to boost Route 96 which is left all by itself on Bourke St. Benefits include simplicity, directness and an extension of Night Network to more CBD streets. Disbenefits include a greater need to change for some trips and having only one tram route on Bourke St (unless another route is shifted).
 

The other approach could be to simply extend Route 30 to Waterfront City. That leaves a more complex network overall that, in having more routes, is less conducive to frequency increases on each one. Hence my preference is for the more direct and cheaper Route 86 option. 

Both options would partly overlap with the rerouted 12. However they would provide La Trobe St with a frequent connection from Flagstaff, Melbourne Central and State Library (a Metro Tunnel station). Either would end the situation where Waterfront City has only limited access to most of La Trobe St and the abovementioned stations.   

- Reroute bus 220 off Footscray Rd to run via Waterfront Way. This would provide better connections to Waterfront City's attractions and a direct route from Footscray. Route 220 is already a frequent service with long operating hours. 

3. 10 minute maximum waits on all tram routes

Currently a lot of tram routes, particularly on St Kilda Rd, operate every 12 minutes interpeak. Route 82 at Footscray operates every 15 minutes (recently improved from every 20 minutes). Sunday morning and evening frequencies are often every 30 minutes. These frequencies are low by world standards and are not quite turn-up-and-go, especially for shorter trips and those involving a connection. Also we've been letting tram frequency decline and stagnate for decades, despite rising population density in areas around them. 

Unless services are very frequent a network coordination framework should specify harmonised frequencies between modes to provide consistent connection times. The rail network has been moving towards an all-week 10 minute daytime/20 minute evening pattern. Timetables associated with the Metro Tunnel will likely accelerate this trend. Trams need to follow with everything operating every 10 minutes or better during the day and never worse than 20 minutes at other times. 

Such frequency increases may require more driver hours with existing routes and speeds. Although the 82 tram got its recent off-peak upgrade mainly by redistributing trips. On other parts of the network there may be scope to reduce overlaps while still serving every stop.  This is particularly so for some St Kilda Rd routes as mapped below. 

Even if you could only remove one route (or substantially shorten two) you might save enough service kilometres to allow several routes to be boosted from every 12 to every 10 minutes. There will be 'swings and roundabouts' but overall most passengers should gain.  


The south-east tram network has an inconsistency based around the weekday Route 3 and weekend 3a (via St Kilda beach). Anzac Station may prove an attractive transfer point for people from the St Kilda beach area. Thus there might be merit in running 3a as the full time route seven days. This then gives a route very similar to most of the 16. Below shows a simplified network based on a new more frequent Route 3 to replace the 16. Route 67 would also need higher frequency to at least partly offset the reduced weekday frequency on a 1km segment of St Kilda Rd. 


The main 'loose end' is what to do with the remaining north-south part of the 16 from Kew. The map has an arrow going to Caulfield. That is indirect. But in the absence of a tram to Camberwell it does at least connect to Glenferrie and further north. Another approach could be to extend it due south to replace the 64 at East Brighton. This leaves the question of what to do with the remainder of the 64 since this overlaps much of the 5 all the way to Melbourne University. 

4. Short shuttle tram trips 

Most tram timetables have all their trips going all their way. That is the same frequency of service is provided at all points along the route. However ridership may vary significantly along it. For example  outer sections of routes might only have five people on board while inner sections may be heavily loaded. One way to accommodate this is to arrange trips so that some start part way along a route. For instance instead of an 8 minute service along the route's full length, there might be a 10 minute service on the outer portion and a more intensive 5 minute service on the inner 'short shuttle' section of the route. 

This could be an example of a 'greater good' change where capacity is delivered where needed but outer area frequency is still reasonable. It is likely only practical on our longer routes only. Active management of dispatches to ensure even spacing and/or loading may be desirable. Also there may be legibility issues where passengers expect trips to go all the way to the terminus but are disappointed when they don't. The least troubling time to introduce this is likely in the morning peak on inbound trips as a relatively small number will be travelling on outbound counter peak trips that stop early and turn back to form a peak direction trip. Again this was proposed by Infrastructure Victoria and supported in the state government response.    

Maybe 10 or 20 years ago another tram shuttle concept was floated for the St Kilda Rd/Swanston St corridor. This was to have a faster and higher capacity spine with larger trams with local trams from the east-west streets feeding in to it, likely requiring a transfer. This received a poor public reception and there seems to have been little mention since. And with the Metro Tunnel providing relief it is arguably less necessary anyway. Hence it will not be recommended here.

5. New CBD area tram links

A tram extension to Arden was recommended by Infrastructure Victoria. The state government didn't support it but didn't rule it out either. However planning needs start now if it's to have a chance of it operating when Metro Tunnel services commence.


IV's proposal for an Arden station tram makes a lot of sense. One up Spencer St involves barely 1.5km of track each way. It would support an already dense (but poorly served) precinct and foster emerging development around Arden. It would connect in to Southern Cross Station, a trip not possibly by rail except via significant back-tracking. Arden would also provide the logical terminus for one of the trams from the south rerouted from Swanston to Spencer St. 

Also mapped is an east-west route from North Melbourne via Victoria St. This overcomes a network limitation of north-south travel being easy but east-west access being difficult. A little like Park St in the south, significant parts of Victoria St already have trams. The concept of a more complete tram grid in this part of Melbourne is nearly a century old, having featured in the 1923 General Scheme


The City of Melbourne advocates aspects of both the above and other tram extensions in its 2030 Transport Strategy

6. New and revised CBD area bus links

Even if you didn't want to do much with buses, there are some changes should be made as the Metro Tunnel will make some CBD area routes redundant or over-serviced. As with cascading benefits for trams there are similar gains for buses with some new or joined corridors possible. Potential bus network reforms include: 

- Delete Route 403 between Footscray and Parkville as its function will be entirely performed by the Metro Tunnel. Resources to be transferred to a new 403 (see later). 

- Delete Route 401 Melbourne University Shuttle as its very high frequency is no longer needed with Metro Tunnel trains available at Parkville. Replace with extended Route 202 to provide a service every 10 minutes between North Melbourne and Victoria Park. Savings could be put towards operating weekend services on Route 202 and/or the new Arden feeder bus mentioned later.  

- Extend operating hours for Bus Route 402 to provide a more useful connection to new Parkville Station. Consider extending to Jolimont or Parliament to provide stronger eastern terminus. 

- Simplify the complex Route 546 in Parkville area with a consistent route. Extend operating hours and upgrade to 7 day service to serve a densely populated but poorly serviced precinct at Alphington.  

- Boost the currently hourly Route 505's frequency to provide a better feeder to Parkville Station. A little money for this was provided in the 2021 state budget but a still higher frequency would be welcome. 

- A new Arden Station feeder bus route to serve north-western areas close but inaccessible to it. Noteworthy destinations and origins include dense housing near Maribyrnong, Highpoint Shopping Centre, public housing near Ascot Vale and Flemington (including that funded for redevelopment), and stations at Newmarket and Flemington Bridge. It could terminate at either Arden or North Melbourne. An option exists to extend it to Southern Cross until there is a Spencer St tram extension in the area. Bus resources could come from deleted or reduced Melbourne University routes mentioned above.  


The new 403 bus serves several public transport 'black holes', areas of high social need and major destinations or interchange points. It connects with 3 (and possibly 5) railway stations. Consequently it should be a good patronage performer suitable for many trips all week. Hence at least a moderate frequency is suggested with a seven day service. The map shows a recycled route number but it is the most suitable for the area given surrounding routes. The main objection to it may be that too much of it duplicates Tram 57's coverage and that if justified at all it should operate at a lower 'coverage route' style frequency such as every 30 or 40 minutes. 

- Other CBD bus route reforms. These are less tied to the Metro Tunnel so I won't go into much detail here. A few examples include efficiencies for Transdev's Fishermans Bend routes (not all need to go beyond Southern Cross Station to Queen Vic Markets), longer hours on the 605 and reform of the West Gate Bridge 232.  

Conclusion

Some central area tram and bus network reforms have been described to complement the Metro Tunnel. The overall effect is to provide better service to inner growth areas including some where development has outpaced service. Special attention has been given to the north-west of the CBD and the Arden precinct. However the network should provide benefits as far out as Maribyrnong and St Kilda. Like with buses, tram reform was proposed by Infrastructure Victoria and supported by the state government. Add them to the network revamps in Part 1 and 2 for the other parts of the Metro Tunnel lines and you would have a transformed network that delivers much of the Double Frequency On Everything plan.

Do you think these are good reforms or could I have gone further? Are some thing ill-advised or are some things missed? Comments are appreciated and can be left below. 

See other Building Melbourne's Useful Network items here

PS: This is the last of a three part series. Part One, covering reformed buses in the area from Watergardens to Footscray, is hereAnd Part Two, describing a similar revamped network between Caulfield and Dandenong is here.  

2 comments:

Michael Angelico said...

Excellent analysis as always.

In general I'm not a fan of indulging the understandable but harmful desire for a single seat journey, so I lean towards your more radical approach to route 86. But it can't be denied that changing trams in the city is a pain, because you arrive at a stop, see the tram just up the street, but still have to wait for ages before it arrives. Traffic light priority would help a lot I think.

Anonymous said...

Oh bring on your new route 403! Highpoint isn't far from me in North Melbourne but the 57 tram drops me at the bottom of a steep hill that I don't like tackling in my elderly enfeebled state. I usually drive, especially if I'm going to a film there, just because the connections are not very good. It has taken me an hour on the 82 tram and the 402 bus.
An old grevience of mine is that the current 401 & 403 buses have no stop in North Melbourne and the station is a fair walk from residences there.
Likewise, when I worked in the city near Spencer Street it was easier to walk than to use any of the public transport options.