Submission to the Rowville Rail Study
.The following is the submission I made to the Rowville Rail Study prior to creating the UrbanMelbourne.info site and was originally posted on my personal blog - as it forms much of the basis in my belief that any new rail lines built in Melbourne should move away from the way we've built them for the past 150 years.
The call for public submissions had a questionnaire - answered first - and a section for "any other comments" which I took the opportunity to link to the blog post so I could explain in somewhat more detail.
Begin submission:
Question 1. In general, to what extent do you agree or disagree with the overall findings and recommendations in the Rowville Rail Study Draft Report?
Answer: Neither Agree or Disagree
Comment: I broadly agree with the general vision in the document - that is connecting this part of the metropolitan region to the rail network in some shape or form, however I overwhelmingly disagree with the central thesis with regards to design characteristics specifically focusing on making the line compatible with the existing network.
Question 2. With reference to the findings and recommendations in the Draft Report please highlight or add, in order of priority, your view of the top three potential benefits and opportunities in relation to the Rowville rail line.
Answer:
1. Monash University rail access and the potential to further diversify development on university land and in the immediate vicinity (more high-density residential, commercial and university-related buildings), further expand the local bus hub
2. Mulgrave rail access and the potential to further develop the land into more high-density office/employment sectors, strengthens the Springvale Road smart bus and local bus routes
3. Waverley Park/Rowville rail access and the potential to further develop the land uses around these stations an, strengthens the Stud Road Smart Bus and has very large potential to turn Stud Park/Rowville into a major local bus hub for the region.
Question 3. With reference to the findings and recommendations in the Draft Report please highlight or add, in order of priority, your top three concerns or issues you may have in relation to the Rowville rail line.
Answer:
1. The idea that all other options other than making the line compatible with the existing broad gauge rail network is short-sighted and disappointing.
2. Utilising existing rail technology and standards will make the line too expensive relegating the project to the "too hard" basket for any state government of any political persuasion.
3. The planned service frequency of a train every 10-15 minutes is inflexible it is reliant on the Dandenong corridor having available paths, the frequency is low, as per the reports design characteristics - existing large rolling stock will be used and there appears to be no thought into marrying up the Doncaster Rail study and Rowville study.
Question 4. What are your thoughts and observations on each of the proposed station locations and designs? - Refer to Chapter 6.
Huntingdale Station:
Answer:
Agree that there should be a connecting station at Huntingdale, however the design as depicted in the report appears to only be necessary for existing train dimensions. Refer to comments below
Monash University Station:
Answer:
The location as depicted appears to be in the best location, it should work exceptionally well integrating with the existing bus hub, however, as per Huntingdale and all other stations as outlined in the report - there will be a significant amount of expensive over-building owing to the reliance on existing heavy-rail technology and rolling stock.
Mulgrave Station:
Answer:
The location is appropriate and should provide excellent connectivity with the Smart Bus on Springvale road and located in an area which has enormous scope for further development on existing properties and the potential for rezoning and reusing the land in the immediate vicinity. The animation is good to show how absurdly long the station infrastructure needs to be (for a much larger cost than what would be required if a separate rail standard were chosen, see comment below) and how expensive on a per-station basis how much capital will be required to serve, in the grand scheme of things, a low patronage stub line.
Waverley Park Station:
Answer:
The location of this station should be near the intersection of Jells Road and Wellington road given that there is existing housing to the North and large employment sectors to the west of the Waverley Park area. Placing this section of the track and station underground seems to be over-kill and elevated track and stations would suffice.
Rowville Station:
Answer:
I believe Study Park SC is the more appropriate location for the terminus at this stage - it's a natural hub for allowing other modes of Public Transport to feed the line and similarly it is a patronage generator in its own right for people travelling further up the line. No need for an expensive underground station - there is large swaths of car parking fronting Stud Road and an elevated station would have less effect on the surrounding area (specifically with regards to the existing road network) than an at-grade station and an underground station appears to be more about exuberance than functionality.
Question 5. Which of the two proposed Rowville station location options do you prefer? - Refer to pages 46–48.
Answer:
Stud Park (prefer to previous answer) - the Wellington Road/Stud Road intersection would make the proposed service a "line to nowhere" with no room for all associated rail infrastructure - i.e mode change to buses through a proper bus terminal, commuter parking etc.
Question 6. In order to accommodate geographical issues and community feedback, the Draft Report has recommended a rail alignment (route) design that includes a combination of tunnelling and above ground rail on viaduct. What are your thoughts and observations on this proposed rail design?
Answer:
Given 250m long stations, underground, would cost well in excess of $100million on a station by station basis, I do not believe underground sections from Huntingdale to Monash, and Waverley Park to Rowville are warranted, with the possible exception of Waverley Park given the grades to the east. Otherwise, cheaper, elevated track with no catenary and smaller 100m long stations utilising Hyundai Rotem or Bombardier ART vehicles is the cheaper and more operational efficient solution.
Question 7. Do you have any further comments?
Answer:
Many.
1. Looking at just this corridor in isolation is a poor use of public funds. For decades transportation planning in Roads has always looked at the metropolitan context rather than just the local context like this study does - take for example Eastlink, a large private road providing an orbital around right through the Eastern side of the metropolitan area.
This study and the underway Doncaster Rail study should be combined because I believe there is a more prudent solution to tackling multiple goals:
- providing rail access to areas that currently don't have them.
- providing a platform to link multiple trunk rail lines outside of the inner city alleviating congestion throughout the middle ring of suburbs and focus further development in specific pockets of the metro area which would be closer to people who live or work outside of the inner city.
- link multiple, already large, patronage generators - Monash University & Mulgrave high-tech precinct, Deakin University, Holmesglen Institute, Chadstone SC, Westfield Doncaster SC and Stud Park SC.
2. The design characteristics are akin to 1950s thinking - the focus seems to be just on the needs of the Wellington Road corridor and effectively end at Huntingdale station, ignoring many millions more people in the Eastern half of the metropolitan area.
Many cities around the world which were unlucky enough to not have a large arterial rail network built a century ago, have been looking for more cost effective ways to provide mass transit to their respective cities and the most pertinent example of this is Vancouver.
The Vancouver Skytrain network has been in operation since the 80s and at the highest level, the differences between it and the current rail network in operation in Melbourne are:
- Smaller Driver-less remotely-controlled trains
- No set track segments - trains do not pass signals which lock when a train is in the segments - the trains having a roaming block / buffer which separates and automatically slows/speeds a train up if two trains are getting too close, this is a key point as it allows the authority running the system much greater flexibility to meet supply with demand at very high frequencies.
- Run at very high frequencies - true metro style, theoretically limits in parts of the Vancouver network are around the 75 second headway mark (48 trains per hour) which at those frequencies would be overkill for this line, but the same technology that can be cheaply deployed, is far more scalable than the current Melbourne Train system.
- Trains on the original network run in 4 x 13m car configurations which carry 80 people (36 seated, 44 standing) per car.
- Trains on the newest Canada Line (the most relevant to this submission) run in 2 x 21m car configurations which carry 167 people (44 seated, 123 standing) per car (334 total - with a crush capacity up to 400 people per 2 car train).
The Canada Line trains are similar in width to one of the three main types of trains running on the Melbourne network at present, however are 3.6m in height - compared to a height of 4.2m for an X'Trapolis class train at present - this height difference is explained by the lack of pantographs above Canada Line trains as they derive power from a 3rd rail setup. Melbourne trains require even more room above the roof of the trains to cater for overhead power supply.
On a train to train comparison in dimensions, X'Trapolis/Comeng/Siemens trains require larger diameter tunnel bores thus increasing underground/tunnelling costs, yet the Canada Line trains carry a comparable load owing to less seats installed compared to a traditional longer-distance train in operation in Melbourne right now.
And this is another critical point I would like to make, the design characteristics in the Report assume every train would essentially originate in Rowville and terminate in the inner-city - mixing with the Dandenong line and eventually Frankston line trains further up the line. I note recent capacity improvements with the 3rd track and platform through Westall and the ongoing need to expand the Dandenong corridor by grade separating and potentially quadrupling the Caulfield-Oakleigh-Clayton section, and as mentioned in the report this would be to the Rowville line's benefit - however what is not mentioned is that in an operational sense, utilising tradition trains is still inflexible and requires planning and timetable changes over a period of months to increase (or heaven forbid decrease) capacity.
An operationally independent line utilising technology similar to the proven Vancouver system will provide a far more flexible service to the people using the line as controllers can remotely add trains into service within minutes - not months - when and where it is needed. The original Skytrain network has a series of 3rd tracks in between the two main running lines scattered throughout the network whereby trains can move in and the controller can then instantly "turn" the train to go back in the opposite direction to provide capacity if it is deemed necessary. This is made possible as each station on the network is monitored from the same control centre via direct CCTV links to every station's platform.
Hyundai Rotem versus X'Trapolis class trains
| Hyundai Rotem | Traditional Train |
|
|
|
Train Height | 3.6m | 4.2m + overhead catenary |
| * Canada Line Technical specs | * Avg Melbourne train height |
|
|
|
Tunnel Diameter Requirements (TBM) | 5.3m | 7.6m |
| * Canada Line Technical specs | * City Loop diameter |
|
|
|
Platform Length Requirements | 100m | 160m |
| * 5 car | *Avg Melbourne train platform length |
|
| 250m |
|
| * 9 Car |
|
|
|
Power Supply | 750v DC Third Rail | 1500 V DC overhead Catenary |
| * Canada Line Technical specs | * X'Trapolis Technical Specifications |
|
|
|
Top Speed | 80kph | 130kph |
| * Canada Line Technical specs | * X'Trapolis Technical Specifications |
|
|
|
Train Width | 3m | 3-3.2m |
| * Canada Line Technical specs | * Avg Melbourne train width (technical specs) |
|
|
|
Seated + Standing Capacity | 337 | 800 |
| * 2 Car | * 6 Car |
| 674 | 1200 |
| * 4 Car | * 9 Car |
| 843 |
|
| * 5 Car |
|
|
|
|
Crush Capacity | 400 | 1000 |
| * 2 Car | * 6 Car |
| 800 | 1500 |
| * 4 Car | * 9 Car |
| 1000 |
|
| * 5 Car |
|
|
|
|
Projected Rowville off-peak frequency (minutes) |
| 15 |
|
|
|
Projected Rowville peak frequency (minutes) |
| 10 |
|
|
|
Potential Rowville-Doncaster off-peak frequency (minutes) | 5 |
|
|
|
|
Potential Rowville-Doncaster peak frequency (minutes) | 3.5 |
|
| * Canada Line standard |
|
|
|
|
Maximum Passengers per hour per direction using standard configuration – off-peak | 4044 | 3200 |
|
|
|
Maximum Passengers per hour per direction using standard configuration & crush capacity – peak | 6857 | 6000 |
|
|
|
Theoretical maximum frequency (seconds) | 75 | 120 |
| * Canada Line Technical specs | * City Loop timetable |
|
|
|
Theoretical maximum passenger throughput per hour per direction (standard configuration) | 19200 | 30000 |
|
|
|
Theoretical maximum passenger throughput per hour per direction (expanded configuration) | 48000 | 45000 |
Canada Line trains: 337 standard seating/standing per 2 car train - crush capacity 400.
X'Trapolis: 397 standard seating/standing per 3 car train - crush capacity 500 (1000 in standard 6 car configuration).
Traditional trains in Melbourne run in 6 car configurations and as stated in the draft report, there would be up to a 10 minute service in peak and 15 minute service off peak. This seems to be low given that enormous amounts of capital will be spent on building very long 250m platform stations to cater for future 9 car configuration trains. This is completely unnecessary and is not likely to have an major impact of getting people out of cars and onto the train if they will be waiting up to 15 minutes for a train!
Compare to a Canada Line system: 2 car configurations which measure 41 metres in length, running on 7.5 minute off peak frequencies on a Rowville-Doncaster metro can move a similar amount of people as 4 trains an hour in standard configuration without the need to spend enormous amounts of money on extra long station platforms or employ the double the amount of drivers (as there would be none in an automated system).
Vancouver's standard off-peak frequency on the Canada line is 3.5 minutes - this equates to 17 trains per hour (TPH) carrying 5700 people in each direction per hour and has a maximum frequency of a train every 2 minutes. 5700 People per hour throughput is consistent with the forecast in the draft report in 2041 (refer to 7.2). The location of turn backs between Blackburn and Springvale roads and between Holmesglen and Deakin University make it possible to operate staggered services should travel demand require it (i.e in Semester times, the loads to Monash, Holmesglen and Deakin would be higher than when not in semester, and these two turn backs can hold 1-2 trains ready to turn around quickly and provide extra capacity to these stations if required - the same thing applies to when Chadstone or Westfield Doncaster see higher demand like during the Christmas shopping periods, demand can quickly and efficiency be met with supply with literally the flick of a switch).
Proposal
Refer to the following map.
View Melbourne PT Plan in a larger map
Implementing a driver-less, centrally-controlled metro system from Rowville to Doncaster via Huntingdale, Oakleigh, Chadstone, Holmesglen, Deakin University and Box Hill.
All stations have 100m long platforms to fit 5 car trains - from day 1 standard configuration would be 2 cars, "marrying" 2 x 2 car sets when required and when the line reaches capacity, introduce a 3rd "trailer" car to some 2 car sets (therefore creating 1 x 3 car + 1 x 2 car sets).
The corridor for Rowville is effectively the same as what has been outlined in the Rowville Rail Study draft report - with the exception of it being entirely elevated. Note I am not a civil engineer, and I'll admit I learnt something about the corridor by reviewing the report especially around the Waverley Park area - in that the grades to the east drop quite a long way and there may just be a requirement to have an underground segment in this area - regardless of whether a traditional line is built or a Canada Line system.
The route is also influenced by the recently reported options released for the Doncaster corridor - specifically this is looking at Option 3 in the Herald Sun graphic.
Arriving at Huntingdale from Rowville, the elevated track would split in two across the Dandenong lines, to the south side of North Road, with the up (toward the city) directional track running in between the up and down (away from the city) directional tracks of the Dandenong corridor allowing for the cross-platform transfer between both the Rowville line and Dandenong line trains. The existing platform 1 and 2 at Huntingdale would be retained and a new island platform for the down Rowville and Dandenong trains would be build adjacent to it in the current car park.
This track configuration would continue to Oakleigh, where the existing platform 3 would have another "face" tacked on to cater for the down Rowville line services. Whereafter the Rowville line would enter the tunnel portal and continue on to Chadstone and eventually Doncaster (refer to map).
This Huntingdale and Oakleigh configuration of the stations achieves multiple goals: commuters from Rowville going to the city (or in the other direction) simply need only walk across to the other side of the platform to catch direct city services. And similarly Dandenong Line commuters can do the same at either station to continue on in the direction of Doncaster and/or Rowville - where providing "one-seat" services to the city is not practical (such as with a traditional train service (inter-mixing with Pakenham/Cranbourne/Dandenong services) or where the rail system in use differs to the existing network) - this is the best and most convenient way to allow people from Rowville who are bound for the city, or from the city to Rowville the most seamless journey.
The Oakleigh-Doncaster tunnel sweeps around Chadstone with an underground station under the recently reported redevelopment of the bus hub fronting Dandenong Road, and further sweeps around to Holmesglen to allow Glen Waverley Line connectivity and also providing further access to Holmesglen Institute & surrounds.
The line then runs under Warrigal road to just south of the Burwood Highway where it then turns east to a station located at Deakin University providing connectivity with the #75 tram, the University and the commercial precinct adjacent to the University. From Deakin University, the line turns north and follows Station St to Box Hill, where two new platforms would be built under the current lower-grade station with direct connections to the existing 3 live platforms and continues running to Doncaster with a station at Westfield Doncaster - the Option 3 from the Doncaster Rail study.
Time example
Section | Distance (m) | Avg Speed (kph) | Minutes | Seconds | Avg Dwell Time (sec) |
|
|
|
|
|
|
Rowville – Waverley Park | 5250 | 65 | 4 | 50 | 10 |
Waverley Park – Mulgrave | 2350 | 60 | 2 | 21 | 10 |
Mulgrave – Monash | 2250 | 60 | 2 | 15 | 10 |
Monash – Huntingdale | 2850 | 60 | 2 | 51 | 10 |
Huntingdale – Oakleigh | 1650 | 55 | 1 | 48 | 10 |
Oakleigh – Chadstone | 1500 | 55 | 1 | 38 | 10 |
Chadstone – Holmesglen | 2100 | 60 | 2 | 5 | 10 |
Holmesglen – Deakin | 4050 | 65 | 3 | 44 | 10 |
Deakin – Box Hill | 3500 | 65 | 3 | 13 | 10 |
Box Hill – Doncaster | 4000 | 65 | 3 | 41 | 10 |
Example usage patterns & time taken to complete journey
Total end to end journey time 29 minutes and 45 seconds (includes 10 second stop at each station)
Existing route segment times from the currently available PTV timetables.
Rowville - CBD (Parliament Station) 41 minutes
- Rowville - Oakleigh 14 minutes +
- 5 minute avg wait in peak +
- 22minutes to Parliament ( express Oakleigh-Caulfield-South Yarra-Richmond-Parliament)
Between Monash Campuses (Clayton and Caulfield) 21 minutes
- Monash University - Huntingdale 3 minutes +
- 7 min avg wait off peak +
-11 minutes to Caulfield (stopping all stations )
Doncaster to CBD (Parliament Station) 24 minutes
- Doncaster - Box Hill 4 minutes +
- 3 min avg wait time in peak +
- 17 minutes to Parliament (express Box Hill - Richmond - Parliament)
Chadstone to Berwick 42 minutes
- Chadstone - Huntingdale 4 minutes +
- 7 min avg wait off-peak +
31 minutes to Berwick (stopping all stations).
Box Hill to Dandenong (Major activity centre to major activity centre) 37 minutes
- Box Hill - Oakleigh 12 minutes +
- 5 min avg wait time in peak +
- 20 minutes to Dandenong (stopping all stations)
Chadstone to Kooyong 21 minutes
- Chadstone - Holmesglen 2 minutes +
- 7 min wait off-peak +
- Holmesglen - Kooyong 12 minutes (stopping all stations)
Deakin University to Mitcham 18 minutes
- Deakin University - Box Hill 3 minutes +
- 5 min avg wait time off peak +
- 10 minutes Box Hill - Mitcham (stopping all stations)
Holmesglen to Templestowe 25 minutes
- Holmesglen - Doncaster 10 minutes +
- 5 min avg wait time in Peak +
- Doncaster - Templestowe 10 minutes (902 smartbus to Fitzsimmons Lane/Porter Street)
Chadstone to Vermont South 25 minutes
- Chadstone - Deakin University 6 minutes +
- 5 min avg wait time in off peak +
- Deakin University to Vermont South (#75 Tram - Stop 63 to 75) 14 minutes
Mulgrave to Camberwell 34 minutes
- Mulgrave to Box Hill 18 minutes +
- 3 min avg wait time in peak +
- Box Hill - Camberwell 11 minutes (stopping all stations)
Keysborough to Deakin University 41 minutes
- Keysborough to Mulgrave (902 bus Springvale/Cheltenham Roads to Springvale/Wellington Roads) 21 minutes +
- 5 min avg wait time in off peak +
- Mulgrave to Deakin University 15 minutes
Knox SC to Monash University
- Knox SC - Rowville 10 minutes (901 smart bus Knox SC to Stud Park SC) +
- 5 min avg wait time in peak +
- Rowville - Monash 10 minutes
Infrastructure Examples from Vancouver's Canada Line
King Edward Station - an example of a double stacked station, Track + Platform + lower/upper level access measures 11-12 metres - applicable uses in Rowville-Doncaster metro: Chadstone, Box Hill and Doncaster (refer to map).
Vancouver City Centre Station - an example of an island platform station, wall to wall the station box measures 20 metres and approximately 15 metres deep - applicable uses: Holmesglen and Deakin University.
Marine Drive station - an example of an elevated station, measures 30 metres wide and the track height above ground is 8 metres - applicable uses: Monash University, Mulgrave, Waverley Park & Rowville
Construction examples
Open air dig of station box, TBM tunnels dug from either end - applicable uses: Deakin University (temporarily use the reserve on University grounds to start tunneling north and south).
TBM Method: to Oakleigh.
(Yaletown-Roundhouse Station)
Side by side cut and cover tunneling method - applicable uses Warrigal Road
Double stacked cut and cover method - applicable uses: Station St in Box Hill to Doncaster, in the lead up to Chadstone station under Dandenong Road
The next four videos put all of the above in to perspective - it's a real time video of the whole line from end to end
Video 1, Predominant construction method: TBM, mixture of island and two platform underground stations.
Video 2, Predominant construction method: cut and cover, mixture of island, two platform and double stacked stations.
Video 3, Predominant construction method: cut and cover, elevated, mixture of island and two platform underground and elevated stations.
Video 4, Predominant construction method: elevated and at grade, mixture of island and two platform evelated / at grade stations.
Costs
Capex
On the map I have made the following cost assumptions:
- Whole line from Rowville to Doncaster is double track with 150m long triple track turn back sections located between Blackburn Road and Springvale Road (elevated) and between Holmesglen and Deakin University (underground)
- Underground track (regardless of TBM or cut/cover methods) averages $50million per kilometre - this is solely tunnelling costs.
- At grade track averages $10million per kilometre.
- Elevated track averages $25million per kilometre.
- Station platforms are 100m in length and priced roughly at $50million for an underground station with no other existing platform connectivity (Doncaster, Deakin, Chadstone), $75 million for an underground station which requires links to other platforms (Holmesglen and Box Hill), $25million for an elevated station along the Rowville alignment and the Oakleigh upgrade costing $10million and Huntingdale station expansion $25million.
- $250 million for the property acquisition and construction of a maintenance and control facility either between Oakleigh and Huntingdale or to the south of North Road passed Huntingdale, also includes staff training costs.
- $5million per 2 car trainset - similar to the Canada Line's Rotems or Bombardier's ART trains. Using the Canada Line as an example, it has 20 2 car trains for 24 kilometres of track which sees normal off-peak frequency of 3.5 minutes in the section of track before the line splits to the southern suburb of Richmond and another branch to Vancouver's airport - this Rowville-Doncaster proposal would require 25 of the same train configuration or $125 million in total.
Opex
In 2008, the Skytrain network (before the Canada Line opened) cost $83million CAD for the Millenium and Expo lines (49 kilometres of track) and with the benefit of not having to employ, train and retain drivers, if a similar system were to be built in Melbourne, this would have an added benefit to the ongoing operation of the network.
There are no publicly available figures for the operating of the Canada Line, and going forward, this should be studied as its bound to have a favourable benefit over the construction and operation of a traditional train line to Rowville.
Conclusion
Whilst I broadly agree with the draft report's general premise, the fact that other types of train systems were overlooked is disappointing. Building underground train lines (or any train line) to fit with the existing network is going to always be more expensive to build - and not to mention operate - compared to an automated system as briefly described above. I disclose my cost assumptions are very high level, and if you're an engineer/planner or someone in the industry who is able to provide a more accurate measure of what has been said above as well as located in the map, I want to hear from you!!
Out of the box thinking is required for the future of Melbourne rail travel, and this proposal as stated above is only but one of three potential phases utilising the same efficient, proven and cheaper-to-build-and-operate technology; Phase 2 from Doncaster to the Craigieburn Line via the Bell St corridor and phase 3 from the Craigieburn Line to Melbourne Airport.
Providing an orbital line that links multiple rail lines that exist now and considering the aspiration of Metro Trains Melbourne to move most existing lines to a Turn Up And Go off-peak frequency of 10 minutes - this connectivity provided by the Rowville-Doncaster line and increased frequencies on the Dandenong, Glen Waverley and Ringwood corridors will throw open so many more non-car transport options to a much wider part of the metropolitan area allowing both local councils and the state government to develop their respective activity centres and provide rail access to areas which currently do not have them.
I urge SKM and all study partners to put it to the state government that both Rowville and Doncaster studies be combined and looked at in the overall regional context and not in isolation using traditional systems and technology already in operation in Melbourne.
