RAIL - Back On Track Forum

Background:

Application of a Commuter Railway to Low Density
Peter Martinovich, PTA WA

https://www.bitre.gov.au/sites/default/files/2019-12/2009_infrastructure_colloquium_peter_martinovich.pdf

Quote from: #Metro on August 20, 2022, 11:14:25 AM

Background:

Application of a Commuter Railway to Low Density
Peter Martinovich, PTA WA

https://www.bitre.gov.au/sites/default/files/2019-12/2009_infrastructure_colloquium_peter_martinovich.pdf

This looks like failure to me!!

QuoteThis looks like failure to me!!

Hey Jonno, 

I figured that this would be your reaction.   

I guess I can only ask what would you change?

I totally understand that:
- Using Car Rapid Transport (CRT) to fill about 30-40% of the passenger boardings and
- Bus interchange for the rest
- Negligible walk-up (5% or less)
- Single use (transport, commuting, and nothing else)
- The whole station precinct is hot and barren, with minimal landscaping (if at all).

isn't consistent with the dense walkable philosophy.
In fact, The PTA calls it "Development Oriented Transit" or DOT which is the opposite of TOD.

The PTA's approach is build good PT first and then density will follow if that's what the market wants.

I totally agree it doesn't look pretty, it is literally set within the middle of a 100-km/hour freeway with fast car lanes on either side. You would not want to spend any amount of time lingering around the station for any reason.

That said, it does the job. It maximises patronage in the given urban context to generate very high patronage (~ 20 million passengers trips/year).

The PTA in Perth did a lot of research into this, I believe part of their thinking during the design phase was to survey car number plates at rail stations and then plot the address locations to work out what the catchment area was.

If the catchment areas are 40 km2 by car, that implies that estimate of the driving distance is about 3.5 km to 4 km (based on the area of a circle). This would give a station access time of 8-10 minutes on average, which is consistent with what we know about the station access travel time budget in the GTCE being around ~10 min.

Compared to say the alternative, Adelaide, Perth has done really well to maximise patronage with what it has got. These good outcomes come from extensive research and adaption of concepts.

It does "a" job but should be a high density TOD but who wants to live near a freeway so this just supports further sprawl! 

This is not DOT either. The literature on that says we should be linking transit to land-use planning and access/connections! I personally don't see a difference between a TID and DOT both highlight the need to link transport planning and land-use/ development planing together. It is a symbiotic relationship!

WA successes comes from feeder buses and frequency!

Not sure the solution given the location next to a freeway! The transit is in a location that you rightly highlight no one wants to spend an extra minute in!!

It is certainly not a model to follow when planning from scratch!  Our urban freeways are going to be a planning challenge well into our future but will they still exist in 50 years!! Who knows! They are a planning mistake that delivered outcomes but the full/true cost has never been really analyzed! Could the benefits have been delivered better! Yes as there are enough cities enjoying the benefits of doing things differently!

The WA PTA distinguish TOD and DOT.

TOD is a change in urban form that is short distance acting - it works primarily by increasing density within a 10-minute walk isochrone of a station.

It acts through bringing houses to the station rather than buses or cars. It improves PT patronage indirectly, through co-location.

The PTA research suggests that 10,000 residents are required in the walk-up zone to reach 500 walk in passengers to support a station. They realise they don't have the numbers for that.

DOT is not a change in urban form. It's the transit itself that is changed to cover longer distances by using unusually wide station spacing (up to 6 km), very high average speed, and using buses and P&R to collect passengers.

I think that is the part that you would disagree with most. Because it doesn't need to or seek to make any change to the urban form. It just accepts it and moves from there. It improves transport directly to influence patronage.

Which means that DOT is inconsistent with the whole movement in New Urbanism which makes The Station the centre of activity. DOT doesn't attempt to do that.

In some ways, DOT is like 1960s and 1970s brutalist architecture. The focus is wholly on utility and function. No weight is given to place making, form, or aesthetics.

I think some of us seem to be living in a fantasy world where extensive urban motorways and low-density land uses in most Australian cities either don't exist or can be magicked away overnight.

It is important not to let the perfect be the enemy of the good.

Quote from: SurfRail on August 22, 2022, 09:15:41 AMI think some of us seem to be living in a fantasy world where extensive urban motorways and low-density land uses in most Australian cities either don't exist or can be magicked away overnight.

It is important not to let the perfect be the enemy of the good.

Equally important to not use the imperfect past as a proof point that the perfect cannot be achieved or worse that the imperfect is the right thing to be do in the first place.  There are plenty of examples of cities changing from imperfect to "far better if not perfect".  Australian city and transport planning is soooooo far behind the eight ball that it is internationally embarrassing.  We actually think and argue that the imperfect is the right thing to be doing.   

I don't think Sydney is behind the 8-ball. They have multiple polycentric cities like Parramatta and clusters like Macquarie Park, Homebush, Liverpool etc.
The density profile for Sydney is very different to other Australian cities and consistently high across the urbanised area.

Note: You can see here evidence that Perth density is almost the same as Adelaide and below that of Brisbane when measured using a square-kilometre population weighted density metric. It's more evidence that our PT system is not performing to full potential. Perth is LESS dense than Brisbane!


Source: https://chartingtransport.com/2019/04/21/how-is-density-changing-in-australian-cities-2nd-edition/

On the other hand, Brisbane and Greater SEQ has a major problem. Even if each city becomes dense and walkable (Brisbane, Ipswich, Beenleigh, Gold Coast, Caboolture, Moreton Bay/Redcliffe, Redlands Bay, Sunshine Coast, Springfield, and new cities such as Yarrabilba etc), none of that is going to change the fact that you can't connect those places to each-other through walking, bicycles etc due to the distances involved. Those connections need to be high speed.

And that most of the jobs are in Brisbane CBD (I really think this needs to change!)

You are still going to need express trains all day and buses doing the pickup.

And this also raises follow-on questions like:

- "How Dense is dense enough?"
- "How many people need to migrate into an area to meet this X level of density?" OR
- "How much does the city need to shrink to reach this X level of density?" OR
- "How much time would it take to reach this X level of density?"

If the standard is realistic, it should have clear and definite terms about what the bar or performance standard to meet is.

In precise terms, what was the alternative for Perth if it was to comply with the expectations of New Urbanism for the entire urban area?

Charting Transport Comparison
Source: https://chartingtransport.com/2019/04/21/how-is-density-changing-in-australian-cities-2nd-edition/

Brisbane Density


Perth Density

Perth I think does try more than Queensland.

Urban development 200m from Cockburn Central.
It is next to a freeway, but they have tried to make a walkable dense area:

https://www.google.com/maps/@-32.1262154,115.8557846,3a,56.9y,27.44h,92.79t/data=!3m6!1e1!3m4!1s8IcidlcHkPnZbi_S9Mkrtw!2e0!7i16384!8i8192

https://www.google.com/maps/@-32.1255316,115.8569162,3a,69.5y,94.04h,101.91t/data=!3m6!1e1!3m4!1saimHmIRQxIkgRDxirqrsSw!2e0!7i16384!8i8192

Wellard on the same line, at least they have zoned the land around the station to be a town center with a main street, high density units
https://www.google.com/maps/@-32.2631141,115.8165661,3a,73.8y,125.9h,94.76t/data=!3m6!1e1!3m4!1sVbR3ChMj72HBTSSxHTySTw!2e0!7i16384!8i8192

This is around Joondalup
https://www.google.com/maps/@-31.7431472,115.7698641,3a,87.2y,72.31h,111.4t/data=!3m6!1e1!3m4!1s3tgeK0eyDhMeoKpSS_iiYw!2e0!7i16384!8i8192

Butler is a bit half hearted, theyve tried a bit of a town center but too much parking.

https://www.google.com/maps/@-31.63529,115.6995362,3a,65.4y,133.86h,98.93t/data=!3m6!1e1!3m4!1svS4vpEzYMVFVhHeFfeMiVw!2e0!7i16384!8i8192

Just remember density is not the "holly grail" for active and public transport.  Nor is total population size.  These have generally been used as excuses to NOT prioritise active and public transport.  As Brent Toderian says "Urban sprawl has failed but car-oriented higher density is a colossal failure". Also remember it is costing up to 6x more in tax-payer funding for every trip that is by car that should be by active and public transport. 

Brisbane has massive opportunities to increase density around our rail/bus stations (even the ones next to the freeway) and our commercial/employment centres (aka the 15min Neighbourhood) plus make these areas walkable, safe to cycle (separated bike lanes/livable street designs) low-car neighbourhoods.  This would also build the actual housing mix we would prefer to live in rather than being forced to choose lower-density car-oriented detached housing because "it is more politically acceptable".

With the right BRT-based (stick to the main roads) network redesign, transit-lanes, station spacing/infrastructure. etc even the most car-oriented areas of our city can have far greater access to jobs, housing, employment without having to commute all over the region.

Agree the days of the CBD centralized employment are over (may not be gone but are certainly not the vast majority of jobs) anymore. This is why land-use/urban design is so critical to undoing the car-oriented city of the last 50 years. Actually I think it is more important than the infrastructure in the short term.  We designed a city that can only work if people are driving day in day out.  Trying to switch people to active and public transport without changing the land use/urban design is I believe actually impossible. For the mode to shift so does the city. The good thing is the car-oriented city is not the type of city people actually prefer to live in.  But no one is educating the general public or having the hard and honest conversation (the media certainly don't help) to say "There is a better way".  Not sure if it true beliefs or just politically embarrassing to admit they last 50 years have been a town planning and transport cluster....and I am an ex-town planner.     

I honestly believe there is no need to build rail lines with few/limited stops and massive, expensive park n rides in SEQ particularly within urban areas as they just make the area around the station barren and unsafe plus are a waste of valuable land. This land is prime employment and commercial space that people should be commuting to not just commuting from.

There is of course the need for express patterns and fast rail (even less stops) but not at the expense of walkable, higher density neighborhoods along the route.  We need to service both. We are spending 75%-80% of the transport budget on building more road capacity which just creates even worse congestion. So we can afford to build it. Just need to stop building the wrong infrastructure.  This is why I don't accept active and public transport being treated like a social service that has to constantly make compromises because "we have to be fair to everyone", or " the approach needs to be balanced" or "the budget is limited" or "[insert any excuse used in the last 30 years to not prioritise active/public transport]".                   

QuoteThere is of course the need for express patterns and fast rail (even less stops) but not at the expense of walkable, higher density neighbourhoods along the route.

I'm not sure why the two are being juxtaposed like that. People in Perth move to around the train station because they want the benefit of the fast trip the train actually achieves. In contrast, just putting people near a train station without the requisite services in terms of frequency and speed (e.g. Doomben Line, most other SEQ train/lines/stations) seems pointless. They are going to live at high density and then still drive.

If we have a project that will yield high train patronage but zero density benefits, should we always reject it?

While we can facilitate development, we cannot force people to live next to train stations. Development is generally market/investor funded. So buyers need to want to pay to live next to stations - so what is going to set up that pull factor? It's certainly not going to be the offer of access to all-stops LSR and VLSR rail services that take 2x the time vs going by car. Especially if it is going to be more than ~ 15 km from the CBD.

But we are forcing them to live away from the train station now! The city we have built (enforced) today is not the city the community would prefer to live in!

Our city is not market led it is city plan restricted!

https://www.dplh.wa.gov.au/getmedia/445a07de-bc30-446d-a4e8-17e39431b2a2/FUT_the-housing-we-d-choose-report

I think the PTA did a good job of working out what the minimums were
(Source, Martinovich, WA PTA, https://www.bitre.gov.au/sites/default/files/2019-12/2009_infrastructure_colloquium_peter_martinovich.pdf):

And I do think that density around stations should generally be encouraged as well. That said I would rather have max patronage now rather than make it conditional on waiting for density to take foot. People will be attracted to stations that offer fast rail connections and frequent service, but the rail service itself has to offer competitive journey times vs car.

You cannot view this attachment.


You cannot view this attachment. 

I wanted to model what is needed to generate Perth-style patronage inside a New Urbanism framework. Here is an attempt:

- To generate the say 700 walk ons, you need to have 10,000 people around the station. This represents a mode share of about 7% (RBOT members may want to experiment with higher shares). If you are going to cut out all park and ride, then you have to raise the bar and generate probably another 600 passengers from walk-up.

- This would require 600 pax x (10,000 people/700 walk-ons) = 8571 more people to be inside that station catchment.

- So the total needed to be generated would be 10,000 + 8571 = 18,571 people in a TOD living around the station.

- The NSW Apartment Design Appendix for a mixed used 5-7 storey block development gives an idea of what is required. I've selected Example 09 as a model input because locals generally dislike any sort of tall tower (p.174).

- This sort of development gives 64 apartments for a site area of 2,840 m2. One development can house 64 x 2.53 average household size = 161.92 persons per mixed-use development.

- 18,571 people x (1 development/161.92 persons) = 114.69 mixed use development blocks in and around the train station.

- The total site area would be 114.89 x 2,840 m2 = 325,726 m2.

- The walkup zone (800m radius of a circle) would be 2 km2 or 2,000,000 m2.

- So this would represent a development of about 16.2% of the walk-up zone to generate enough to support the station.

Comments

This is an idealised model, but it gives some idea of what scale of TOD development is required. Others can adapt the numbers to see what effect changing different conditions create.

Just generating minimum patronage would need about 62 of these 5-7 story mixed use developments around the station (assuming the train service is any good vs alternatives).

If you delete the car park, then the rest need to be brought in by buses (how would that be increased?) or have to come from more construction inside the walk-up zone (increase from 62 to 115 mixed use developments around the station).

Reference

The NSW Apartment Design Guide Appendix is here, the model apartment is on p.174

https://www.planning.nsw.gov.au/-/media/Files/DPE/Guidelines/apartment-design-guide-appendices-2015-07.pdf?la=en

Quote- To generate the say 700 walk ons, you need to have 10,000 people around the station. This represents a mode share of about 7% (RBOT members may want to experiment with higher shares). If you are going to cut out all park and ride, then you have to raise the bar and generate probably another 600 passengers from walk-up.

If we are targeting just 7% mode share then we might as well pack up and go home.  If we make these assumptions based on  car-centric design and infrastructure then the majority of the trips are then planned to be by car and then the infrastructure is built first because it is the bigger % and around and around we go.

We need to be planning and creating neighbouhoods and infrastructure that enables/facilitates much higher % of active and public transport, with more local trips (mixed-use/15min Neighbourhood) but also also a mix of commuting directions and purposes.   

As I said above we can't create a mulit-modal city without:

• changing how our cities are designed and built; and
• prioritisng active and public transport infrastructure build

QuoteIf we are targeting just 7% mode share then we might as well pack up and go home.  If we make these assumptions based on  car-centric design and infrastructure then the majority of the trips are then planned to be by car and then the infrastructure is built first because it is the bigger % and around and around we go.

I believe ~ 7% mode share is what Perth actually gets at train stations on this line?

The mode share is influenced by the General Travel Cost Equation - that is how fast the overall journey times are versus the alternatives (e.g. car).

Simply placing 10,000 people in a TOD around a station that might have poor services (e.g. Cleveland Line) will mean that they simply live around a train station and drive IMHO.

Yes but I think what Jonno said is that Perth has pretty decent services on all rail lines, but still only gets 7% mode share.

This is still a bit disappointing, and means more extensive changes are needed in terms of how we actually build our cities to shift this figure.

But as Surfrail said, we have a lot of city that is already built

For example this subdivision is 800m from Wellard station and looks late 80s/early 90s
https://www.google.com/maps/@-32.2647499,115.826481,3a,75y,105.79h,77.27t/data=!3m6!1e1!3m4!1soVaridn-108JuwzDV1Ly2w!2e0!7i16384!8i8192

These areas are probably not going to be knocked down and densified because there is just so much of it.

We can make sure new suburbs are denser/walkable.

But to give a Brisbane example, what should we do with Forest Lake?

Be wary (I certainly am) of modelling as it is this modelling that has seen us develop car-dependent cities and time and time again discount active and public transport prioritisation because the "modelling shows it won't work" or "it has a low rate of return".  The modelling is broken or biased or just plan wrong.   

Yet there are cities all around the world that show prioritising active and public transport (remember we are lightyears away from doing this) does work and they are so much better off. This includes changing the cities design/urban form.  I think we are better off modelling our cities on these cities including the frequency, prioritisation, interconnectivity and fares of services.  Or at least use them to understand why the modelling has been so wrong.

Edit to add thoughts on Gazza's question:

I think the key immediate actions whilst it takes does time for neighbourhoods to become 15Min Neighbourhoods is to:
1. educate the community on the need to change.
2. make cycling safe and separated on roads and turn streets in low-speed/livable whilst minimizing through traffic (low-traffic neighborhoods).  This massively extends the reach of public transport.
3. make public transport fast, direct, prioritised (own-lane) inter-connected and frequent.  Not stopping every 200-400 m but building proper "light-rail like" stations next to the local shops or major intersections.  Assume it is light rail even if it is not.
4. Finally make our commercial and employment centres low-traffic and walkable.  We have just made it far to easy to drive and park. It is a cultural issue not a modelling issue.

I think we also under-estimate the pent up demand for housing choice right across our cities. Our City Plans need to be changed to (1) stop car-centric low and high-density development from being approved in the first place and (2) allowing development to occur as and where it is wanted.  Current the City Plans stop it so "there is no demand clearly". 

QuoteBe wary (I certainly am) of modelling as it is this modelling that has seen us develop car-dependent cities and time and time again discount active and public transport prioritisation because the "modelling shows it won't work" or "it has a low rate of return".  The modelling is broken or biased or just plan wrong. 

Models aren't right or wrong, they are models that are better or worse fit to the observations or axioms.

New Urbanism is a model too that has its own assumptions. Like very high mode share will be generated next to stations if people are placed near them. But in what quantity?

That Perth 7% mode-share is across all work/uni trips, right? Not just for CBD stuff.

We could plonk a big old TOD at Cleveland (to use Metro's example) but anyone who works at, say, Capalaba can't take the train and maybe even isn't taking the bus.

What this really calls for is a cross-tab of census data but what's needed here is sensitive enough that it's never getting released to the public. (Place of residence, place of work, method of journey to work.)

And, of course, a grid network that supports non-CBD trips, and a general setup that supports a lot more cycling for trips under 5 km.

Quote from: aldonius on August 23, 2022, 11:46:12 AMThat Perth 7% mode-share is across all work/uni trips, right? Not just for CBD stuff.

We could plonk a big old TOD at Cleveland (to use Metro's example) but anyone who works at, say, Capalaba can't take the train and maybe even isn't taking the bus.

What this really calls for is a cross-tab of census data but what's needed here is sensitive enough that it's never getting released to the public. (Place of residence, place of work, method of journey to work.)

And, of course, a grid network that supports non-CBD trips, and a general setup that supports a lot more cycling for trips under 5 km.

Think we just said the same thing. 

Well, what are the mode shares in those suburbs that are dense and already meet the "dense enough criteria"? For both inner city and outer?

Are there any suburbs that already meet those ideals and if so, what are their mode shares? The ABS has journey to work data. 

And how dense is dense enough to meet the minimum required to support frequent train service?

See #10 above.

2021 mode share won't be available until October, but here's an indicator of 2016.

Great Map Aldonius. Is there a corresponding one for Perth I wonder?

QuoteSee #10 above.

QuoteWith the right BRT-based (stick to the main roads) network redesign, transit-lanes, station spacing/infrastructure. etc even the most car-oriented areas of our city can have far greater access to jobs, housing, employment without having to commute all over the region.

I actually think this aim is doubtful, there is always going to be a need for regional travel.

Take UQ. It is Brisbane's second largest employer and second largest destination after the CBD.

We cannot expect a UQ-sized university to be within 15 minutes walk of everyone negating the need for regional travel. The 15 minute isochrone is also very small.

You could spend that time walking to the bus stop and waiting for a BUZ (10 min walk + 7.5 min average wait = 17.5 min, is the time budget exhausted). Even the bus to UQ Lakes from the CBD takes 20 minutes with the busway. That is outside the 15 minute isochrone. What if you live on the Northside and you work at UQ?

The same goes for major hospitals. or Government Departments. These are very large employers.

And major shopping centres with specialist retail. 

Then there are industrial workers who work in industrial estates and warehousing/manufacturing plants that are unsuitable for inclusion in mixed use residential developments.

Major transport connections like Brisbane Airport cannot be within 15 minutes either.

Density etc. is fine, but I think squeezing everything or even most things into a 15-minute isochrone is going to be impossible. Particularly for the journey to work, which is a huge generator of trips.

For those, you are going to have to introduce an additional 30-minute or 45 minute isochrone for that to the existing 15-minute walkable neighbourhood one. Otherwise it would be challenging to see how even existing Australian cities would work.

One challenge we will always face is that we can create walkable communities, but jobs will continue to exist in the low density areas outside of these (For example an older suburban shopping plaza surrounded by parking like Mt Ommaney)

A BUZ route going past these places will help, but I'm unsure how you prevent masses of people going from an established low density resi area to an established low density job area.

I was talking about large % people who now regularly commute for their job from Noosa, West of Ipswich and the Gold Coast.  We need to enable regional mobility/access without a doubt for the exact reasons you specify but focusing on regional daily commuting (aka well over an 1hr) is not healthy, adds significantly to cost of living and the unafordabilty of housing nor is it sustainable nor desirable. I guess there will be a group who wish to but will a much more diverse employment base these become the exception not the norm over time.

QuoteI was talking about large % people who now regularly commute for their job from Noosa, West of Ipswich and the Gold Coast.  We need to enable regional mobility/access without a doubt for the exact reasons you specify but focusing on regional daily commuting (aka well over an 1hr) is not healthy, adds significantly to cost of living and the unafordabilty of housing nor is it sustainable nor desirable. I guess there will be a group who wish to but will a much more diverse employment base these become the exception not the norm over time.

But what if you want to go between Brisbane and the GC/Sunshine Coast because you want to go to the beach? Or a theme park? Or you have friends/family there. You can't put that within 15 minute or even 30 minute isochrone.

I don't think regional travel like this can or should be stopped. Short of putting up something like the Berlin wall at the 30-minute or X-minute isochrone.

(but you could charge for it in distance-based tolling/road user charges)

Commute - Definition

• Travel some distance between one's home and place of work on a regular basis.
• A regular journey of some distance to and from one's place of work - "the daily commute"

Quote from: #Metro on August 23, 2022, 17:07:48 PMGreat Map Aldonius. Is there a corresponding one for Perth I wonder?

There is now!

WOW!

Great stuff Aldonius.

Based on this information, what then is a "reasonable" mode share or mode shares for PT for using in a model?

I've done some more lookup, as I'm now curious about more shares for some of the denser cities or towns that are being used as reference models.

Ideally we would have more granular data at a suburb level, particularly for areas say outside of the 15 km zone but inside the 30 km zone.

Mode of transport used by commuters to travel to work in the Netherlands in 2019
https://www.statista.com/statistics/1013713/mode-of-transport-used-to-commute-to-work-in-the-netherlands/

For Netherlands overall it was about PT=22%, which is comparable to more share for inner city Brisbane.

For outer parts of Brisbane, this was about 15-20%. Would be good to see what other cities have. It seems reasonable that 20-25% mode share is the upper ceiling on PT mode share given the state of current services assuming that everybody in greater Brisbane used PT like West End residents do.

High-PT Use Scenario

From the PTA slides earlier in this thread, the WA PTA aimed for 2000 boardings generated from a catchment of 30,000 people. This isn't assuming that everyone is in the walk-up zone but various modes such as also bus, bicycle, drop-off etc.

This gives (2000 boardings/30,000 people x 100 = 6.6% <---- About 7% mode share.

We are now going to tweak the model a bit more and use 25% as the PT mode share, which would be to assume that PT usage was generated at West End Brisbane levels inside our idealised urban catchment.

If we have the following set of assumptions:
- Set 25% PT mode share
- 2500 daily boardings as the minimum for supporting a rail station
- What is the catchment?
- At what density does this have to be?

So 25% = 100 x (2500 boardings/x),
Solve for x = (25%/100) = 2500/x
==> x(0.25) = 2500
==> x = 2500/0.25
==> x = 10,000 people in the catchment. This is a reduction of about 1/3 from 30,000 in the catchment.

In this model, generally, the number of people required in the catchment is simply the minimum required to support the station divided by the mode share.

Comments

The inverse of 1/3 is 3x.
This suggests that we could alternatively densify the catchment by a factor of 3x, either through:

- Adding 5-7 storey density near stations (will be challenging to get 65 x 5-7 storey developments at PT=7% done in the 800m radius walkup zone from the station to generate the minimum to support the station so we should loosen this a bit)
- OR if we allowed single detached homes to allow triplexes on them
- OR if we shrunk the lot size of single detached homes
(or any combination of the above)

Maybe we could get to that 25%. This is consistent with what we know about West End, which has seen massive development.

Most of Brisbane already has PT mode share at or above what is observed in Perth. If the rail timetables were modernised, we should expect the same or better levels of patronage on trains.

And the above model implies that maybe the dense mixed-use neighbourhood might not be the only way. If changed zoning rules to allow triplexes on all properties zoned for single homes, or perhaps shrunk minimum lot sizes (or a combination) that would be sufficient to do the job.   

High-density mixed-used neighbourhoods get a lot of attention because they have good optics, but I wonder if there is a more dull but equally effective (and maybe politically palatable way) to get to this PT=25% mode share. e.g. Replace single homes with duplexes on small lots or triplexes on normal lots.

Not everywhere is suitable for a 5-7 storey mixed use development. That is a very intensive use of land. But a 3-plex everywhere or duplexes on smaller lot size might just work. You would just need a lot of them over a wide area. They are also much less expensive to construct than 5-7 level multi-storey buildings because there is no need for lifts, fire exit stairs, etc. Also potentially less likely to trigger NIMBY.

This above exercise also highlights just how anti-public transport BCC's zoning ordinances against things like townhouses in the suburbs actually are.

Reference
Marinovich (2009) Application of a commuter railway to low density
https://www.bitre.gov.au/sites/default/files/2019-12/2009_infrastructure_colloquium_peter_martinovich.pdf

Please read - https://missingmiddlehousing.com/

And

https://blogisthmus.files.wordpress.com/2013/04/isthmus_gentledensity_sm.pdf

And

https://www.smartcitiesdive.com/ex/sustainablecitiescollective/ten-steps-walkability/96836/

And

https://www.arup.com/perspectives/designing-the-fifteen-minute-neighbourhood

These about building healthy, vibrant, diverse, sustainable neighborhoods that are walkable and safe to cycle!!

This has to go hand in hand with the prioritising of active and public transport or we end up with car-oriented higher density! The colossal failure!

Whilst they have a symbiotic relationship the provision of infrastructure/service and good connectivity, priority and quality of service plays a very big role in changing mode share!!

https://www.ptua.org.au/myths/density/

There is something simple and neat that follows from this discussion:

Q: When should you build and maintain a large car park around a train station (e.g. 500 car spaces)?

A: When the number of passengers brought in by the car park is greater or equal to those that would have walked up to the station if the area within 800 m of the station were developed.

- This is a re-appearance of the same problem we saw with questions such as "When should a bus lane go in?" They all depend on identifying whether the passenger volume is above, at, or below a floating Point X.

- Based on this, many of the stations on the TransPerth network probably do justify large Park and Rides. Particularly if there were no other way to generate sufficient walk-up patronage. As it represents the highest and best use of the land at that point in time.

- As land value increases at some point the balance would switch from TOD < Car Park to TOD > Car Park and the parking can be removed and developed.

- You could allow the land to be bought by a developer at any time on the condition that they provided at least X number of dwellings, or more than that.

Market Sounding Test: The Transit Authority could offer the land to the market every year for development until an offer came in that matched Point X.

Example:
- For a 1000 car park to be converted, the Transit Authority would need to receive an offer from a TOD developer to build at least as many apartments around the station to replace every passenger generated from the existing park and ride.

- We need to estimate the PT mode share for this to work, but that is not too difficult (reasonable band is say PT=5% to say 25% max.)

So, assuming PT=7%:

1000 car parks x 1.2 load factor = 1200 passengers generated by Park & Ride (the 1.2 adjusts for 20% of the cars carrying a passenger)

1200 passengers x (1/2.53 average household size) x (1/PT=0.07) = 6775 dwellings (Low PT use scenario)

- IF there are objections to this PT=7% being "too low" then we can simply increase it to the "correct" level (choose your number):

1200 passengers x (1/2.53 average household size) x (1/PT=0.25) = 1897 dwellings (High PT use scenario)

So if a developer came and offered to build at least 1897 dwellings around our model station, the Transit Authority should probably accept. Note - this does mean that the car park will be removed when it is at full capacity, so I would expect a lot of noise from former car park users.

Why use PT=25% as the high water mark level?

Because that is the highest observable PT mode share for residents of the CBD. CBD residents literally have a (direct) bus, train, or ferry to every suburb in the entire city and can travel in any direction. They have the best frequencies and connections. This is therefore the high water mark for any other test suburb we saturate with the best possible service.

If we ever observed a higher PT= __ % number anywhere in SEQ, we would simply adjust the model up in future iterations to match that.

A point of discussion would be whether we use the mode share just for peak hour, or all day.

QuoteHigh-density mixed-used neighbourhoods get a lot of attention because they have good optics, but I wonder if there is a more dull but equally effective (and maybe politically palatable way) to get to this PT=25% mode share. e.g. Replace single homes with duplexes on small lots or triplexes on normal lots.

Turns out NZ and US cities are doing this - it has a name too - UPZONING.   

Upzoning doesn't insist that you have to redesign every neighbourhood to look like Vancouver/Amsterdam/Copenhagen or even develop inside the 800 m walk-up zone. It just allows you to double or triple the density using a do-minimum approach. Instead of one single family home, you have duplex or a triplex. Or split up an existing single family home, they are that big.

Do that and also throw in more BUZ routes and add more trains on existing lines and we have a practical solution IMHO.

Solution = Strategy A (Density around stations) + Strategy B (Upzone everywhere else) + Strategy C (BUZ & TUZ)

QuoteTo help address a housing shortage, Minneapolis became the first large American city to end single-family zoning, the rules that restrict certain neighborhoods to single-family homes. Now, buildings with up to three units can be built on any residential lot. Leaders hope this, and other plans, will add new units, create density and remedy segregation. NewsHour Weekend's Megan Thompson reports.

QuoteDensity

The image below shows two towns and a transit line running through them.  The transit line in the two images is identical, so it has the same cost.  But in the first town, the line is available to twice as many people, because the town is twice as dense.  That means that if everybody in both towns had the same propensity to use transit, ridership would be twice as high on the first route as on the second.

Source: https://humantransit.org/basics/the-transit-ridership-recipe#density

I have have a copy of the following Conference Paper by M. Peter Martinovich from the WA PTA.  
Unfortunately, I cannot share it as it is paywalled (though it may be possible to get it directly from the author).

CORE 2008: Rail; The Core of Integrated Transport Conference Paper 01 January 2008
The Integration of Rail Transit and Land Use in Western Australia
Contributors: Peter Martinovich

https://search.informit.org/doi/epdf/10.3316/informit.563119378388351

QuoteThe traditional understanding of mass transit rail is its application to serve high density settlement. This paper is about the application of mass / rapid transit rail to low density settlement in a city of just over 1.5 million people and rapidly expanding over a 135 kilometre long, narrow, corridor...

The paper discusses the Importance of the Master Planning process, which quantifies the demand and then follows a logical process of scope definition to reduce uncertainty and give confidence to the cost estimates that result. It also addresses the issue of how Transit Oriented Development relates to Perth.

Some key points in the paper:
- Originally planners et al wanted a bus and busway system. They actually built one from the CBD to Murdoch station.
- Density was very low and car use high so LRT and bus was thought to be more appropriate
- The original Mandurah line proposals were going to send rail on an indirect route (a) either via the Fremantle line or (b) via Thornlie-Armadale and then into the CBD
- Transport Department said NO to use of Kwinana freeway median into Perth as 'Not enough space' (closer analysis showned that it was possible)

Quote"The Direct Route including tunnelling through Perth was soon subjected to sustained opposition by city fathers and disparate groups including the major professional bodies.

There were claims backed by engineers that the freeway median was too narrow and that the major road bridges along the route could not support trains. Fears were raised about the impact on the river foreshore in Perth, on engineering feasibility, on city building infrastructure, impact on commerce along William Street during construction, safety, transport and land use outcomes."

Motorised access to stations using buses and P&R are essential for high patronage in a low-density setting.

Quote"Surveys show that up to 91% of commuters come from beyond the walking radius by car or bus."

Using P&R makes sense when you weigh up the travel time comparisons. A bus would take 2x longer, and cycling even longer than that. So it is not simply a case of deleting the P&R car park and then hoping people will switch to cycling to the station - that would simply increase their access time beyond their 10 minute time budget at least by 2x, and they would be lost as a passenger.

Another issue here is that TOD proponents tend to treat all car use as equally bad. So a car trip that is formed as a connection to a rail P&R and a car trip that is formed as a direct to the CBD down a congested freeway are treated the same. This view lacks nuance - a trip to the local train station isn't as bad as one going to the CBD and the VKT generated, CO2, fuel used, road infrastructure required, and congestion created by such a short connecting trip will also be far less than one that went all the way to the CBD.

QuoteWhat stands out from the experience IS that the short, private car trip to reach transit is highly valued. The value of time especially in the morning, the car's flexibility. convenience and availability on call, are the positive attributes that make access to stations by car attractive

Figure 12 typically shows that the journey from front door to platform is likely to be at least twice as long for the bus, compared to parking at the station

Background to the Rockingham Light Rail Proposal