Following on from yesterday’s post about the question of when we might need to think about rail to Auckland’s North Shore, there’s also an interesting question of where it should go and what form it should take. The Passenger Transport Report that accompanies the Harbour Crossing documents looks at this issue in some detail.
Effectively, there are two options when it comes to answering the question of ‘where to put a railway line?’ You can either follow the existing busway alignment, or you can go along a different alignment. Of course there are a number of possible options for a different alignment (east of the busway, west of the busway or a mixture). There’s also a question of whether we should use light-rail or heavy-rail. The report focuses more on the first issue than the second, although my personal opinion is that it would be rather pointless to upgrade the busway to light-rail – as you wouldn’t increase capacity much at all (in which case why bother?) So in my opinion it’s heavy rail or keep the busway as is.
One thing that’s useful to look at first, before deciding on our preferred alignment, is the interaction between land-use patterns and our possible options. Obviously a big difference between busways and railways is the “feeder” issue: with a busway you can have a bus running along suburban streets before joining the busway itself – giving people a one-seat ride from their home to their destination. Something like what’s shown in the diagram below:
In contrast, obviously with a railway line you need to do things a bit differently – and people need to transfer from the feeder bus onto the main trunk line:
In a way, the success of a rail system is likely to be dependent on two factors. The first being whether you can build the system in such a location that will make it possible for a large number of people to live around the station (and thereby maximise your ‘walk-up’ catchment). The second factor is whether you can build you system in such a way that encourages people to use feeder buses – meaning you’re less reliant on the ‘walk-up’ catchment and the feeder buses effectively act like extensions to the railway line. I would suspect that high frequencies (of both feeders and the rail service), integrated ticketing & fares, a physically easy transfer and a rail speed high enough to ‘make-up’ time lost waiting for your service are critical to making the feeder system work.
There are a variety of international examples of either type. Perth has been very successful in attracting passengers to its Mandurah and Joondalup lines through feeder buses: creating successful railway lines through areas of residential density thought far too low to support rail. Similarly, around 70% of people using the Toronto Subway system arrive at their station by bus. However, in other cities you don’t see nearly the same results – a look at PT modeshare around Sydney shows that rail use drops off dramatically once people get further than 1-2 kilometres from the nearest train station:
Blue shows rail, green bus and pink ferries. The darker the colour the higher the modeshare and the larger the circle the greater the number of people. Overall, what the map seems to show is that high rail use is very much concentrated closely to the corridors: once you get much further out from them the use of rail drops off fairly dramatically.
Obviously in an ideal world we would want a North Shore Railway Line to serve the denser parts of the North Shore, or at least serve areas where higher density development was possible. Right next to a multi-lane motorway doesn’t exactly fit that description particularly well, and as the map below shows, the motorway/busway corridor is a bit of a “hole” in terms of residential densities on the North Shore:
The highest density areas on the North Shore seem to be set back a kilometre or two from the motorway, both to the east and west. The whole map is actually quite interesting to look at, and once again in my opinion helps to show that Auckland actually isn’t a low-density city, it’s just a city that doesn’t have too many very high density areas. A further graph, comparing Auckland with Wellington and a number of Australian cities, seems to confirm this observation:
So when compared to Brisbane and Perth it would seem that Auckland has more potential for rail to be successful in normal suburban areas: because the density ‘bulge’ is around 30-50 people per hectare whereas Perth and Brisbane have their ‘bulges’ at densities of between 10 and 30 people per hectare.
But anyway, if we get back to the North Shore issue, I can certainly see some argument for wanting to find an alignment different to that of the busway, in order to put the railway link through: to connect up areas with the highest density and the most likelihood of having higher densities in the future. This seems to be the preferred option of the report, which suggests the following alignment:
The big problem with this alignment is the cost. The whole thing would need to be tunnelled, as there is no existing corridor for it to pass through. Add that in with a fairly grand plan to connect with the existing rail system at the city end, and you have one hell of an expensive project:
Yes, that is a total of $11.5 billion. Personally I think the figure is bloated enormously as it would be totally unnecessary to spend $307 million on every underground station (my reading of the CBD Tunnel business case is that even the deep K Road station’s cost was ‘only’ around half this amount). Plus I think a Hospital and Symonds Street station is a completely unnecessary way of connecting it in with the rest of the city (my preferred option is shown in more detail here).
The huge cost, in my mind, completely rules out any alignment for rail on the North Shore other than up the busway. Furthermore, I can’t actually see the alignment proposed above having too many significant advantages compared to the busway as it follows it fairly closely most of the way – only branching off at Takapuna and at Mairangi Bay really. So much extra cost for so little relative gain.
Now that’s not to say that upgrading the busway would be a piece of cake, and the potential difficulties of this are outlined in the report: 
I can understand there would be costs associated with fixing vertical alignments (probably a tunnel underneath Sunset Road ridge would be needed) and also extending the stations would also cost a bit of money. It’s likely the Tristram Ave overbridge would need to be strengthened or rebuilt and there would need to be widening (just take a motorway lane I reckon!) But the cost of all of this is likely to be very little compared to the $11 billion fully underground alternative. The real question is whether we want our rapid transit corridor to be right next to a motorway, and therefore always likely to be more reliant on feeder buses rather than a walk-up catchment?
Ultimately this comes down to the question that we started our previous post with, under what circumstances would the Northern Busway no longer be enough? Essentially, in my opinion the reason to build rail on the North Shore will be to solve capacity problems: either in the CBD or on the busway itself. It won’t be as much of a ‘place-shaping’ exercise as other railway projects, like the CBD Tunnel and potentially rail to the Airport. But that’s OK, Perth and Toronto show that a railway line based on feeder buses can be successful – particularly if it’s frequent and fast. The real benefits of upgrading the busway to rail will be felt in the city centre, through a reduction in buses, for those on the North Shore who can hook into the rest of the rail network and are able to get on a rapid transit service that isn’t clogged (as the busway could be by 2030-2040) and for the people of Auckland and New Zealand who have avoided spending $5 billion on a road tunnel, or $4 billion on a bridge that’s ruined the harbour.
So, for me, it’s heavy rail up the busway.
Processing...
I saw a diagram similar to the Sydney one above a few years ago- showing PT mode-share in different areas of Auckland, colour coded for the % of trips. I cant seem to find it now, but it was fascinating to see where people use PT (all central suburbs, and those near the RTN), and where people dont (Howick, Whangaparoa, far west Auckland, Otara etc. I think it was on a green Party blog actually. Does anyone have a copy?
Erm, an interesting set of conclusions.
It says new longer stations would be required. But the existing station sites are over 200m long. Surely this doesn’t mean new stations are required, it just means that rail platforms would need to be built which is pretty obvious.
A new rail alignment would be required from Constellation Dr… well duh, obviously! Any option would require an new alignment north of Constellation Dr because there isn’t anything there already.
I don’t agree with the widening point, the busway is 12m wide, that’s as wide as either of Perth’s two new lines in motorway corridors and the use much the same track and loading gauges. So why would ours need a corridor 15m wide?
The point about population and employment densities is a complete red herring too. Nowhere on the Shore (and hardly anywhere in Auckland for that matter) is going to give sufficient densities to support walk up stations. We are talking about a line fed by buses and park and ride here regardless of the route. If they think moving Constellation station to an underground location at Windsor Park will change anything they are surely mistaken!
It seems like they have gone out to design a metro line, when they should have been designing the mandurah line instead.
I agree Nick this sounds like a load of expletive. The Western Line outside my house would be no more than ten metres. I’ll measure it and find out, I’ll be generous and measure from barrier to barrier. I think they’ve gone to a station, measured the width of the corridor at 15 metres and then applied that to the whole corridor.
Also they are already looking at extending the busway to Albany. So *duh* the work on the new alignment will already be completed and hopefully built by the time we look at this.
One just hopes if they do extend the busway to Albany they design the extension to heavy rail geometry.
It seems that they are assuming we must have both busway and rail line…. can’t see what else they’re getting at? Unless it’s just Road-heads charged with coming up with reasons for dismissing rail…
If you read between the lines they have basically decided it is too hard to manage the conversion (there were some reports of “shutting down the entire busway for two years”) so the only option is a new alignment which therefore has to be fully tunnelled.
It’s just crap really. Obviously CBD to Akoranga and Constellation to Albany would be entirely new sections, so no conflict there, while the rest could be modified in stages with only one section shut down at a time. You basically build the harbour tunnel to Akoranga, test the line and the vehicles, then open it with Akoranga as the temporary bus to train terminus. Then work your way back from there, i.e. close and rebuild Smales to Akoranga, with buses using the busway to Smales then a diversion around the work site to join the trains at Akoranga. Then you open the rail line to Smales and close the section to Sunnynook, etc.
Arranging temporary busway routes on the motorway, motorway shoulders or local streets wouldn’t be so hard. If they can replace the Newmarket viaduct and double the amount of Spaghetti in the CMJ without closing them down they can easily do this.
I have to agree with you Nick R.
Perth’s Mandurah line goes down the Kwinana Freeway and previously replaces the median busway they had there.
The rail line runs in previous busway lanes, and as I understand it, Canning Bridge Bus station was converted to a train station.
I am scratching my head too.
A backgrounder on the Ottawa Transitway to Light Rail
http://homepages.ucalgary.ca/~dpjames/mdp/david_james-mdp-final.pdf
Cheers Bris, section 5.7.1 and 2 describes basically the ‘incremental’ upgrade I was trying to describe above.
“The incremental conversion strategy would see each section of transitway (essentially
station to station) converted to a fully operational state before moving on to convert the
next section. In other words, steps 3 through 6 would follow step 2 and then the cycle
would be repeated to the next station and so on until the transitway was fully converted.
Of note is that the
incremental conversion strategy requires that each station become a temporary modal
transfer station at some point in time.”
The last point is no problem. All the busway stations (excluding Sunnynook) are already modal transfer stations between the NEX buses on the mainline platforms and the local buses on the local platforms.
SPEED!
If you put on a good, fast public transport service, people will fall over themselves as they scramble to access it.
Don’t worry if the busway or railway isn’t passing through high density. The Brisbane busway doesn’t do this either, and in fact 80% of people who use the Brisbane SE busway don’t set foot in a busway station. They get brought in by boring old bus stops in the suburbs.
In my view, the quality Public Transport has to come first, then the urbanism can come next automatically as developers see that the service is fast; not the other way around. Placing rail down the busway is also cheaper as this can use the grade separation of the busway rather than start new tunnels, demolish homes, move services etc from scratch. Anything else needs huge earthworks etc so that’s going to be $$$.
As for the city density profiles, I’m not yet convinced anything meaningful can be said from reading them. Both Brisbane and Perth have peak density below that of Auckland, and yet both cities carry around five times the numbers of passengers that Auckland rail does and both their mode shares for journey to work are much higher than Auckland’s.
Heavy rail isn’t the only option down the busway. Larger buses might be an interim solution. Light Metro (Vancouver) can go to 30 000 pphd, and true metro can go higher than that. I see metro systems as just ‘higher capacity rail’ and they might be something to look at.
What’s the ticketing integration like in Brisbane and Perth? A huge impediment to PT use in Auckland is the zero-integration model we have, which drives up cost and inconvenience to levels that make it impractical for many potential users. If you have to bus-train-bus and pay a separate fare (or over $200/month) for each step, why bother when it’s probably cheaper to drive and definitely more convenient.
We also don’t have integration of services, so buses aren’t timed with trains, etc. Again, a huge headache, and it even applies to transfers within the same mode. I’m not sure what it’s like now but apparently there was a point fairly recently where the timing between services on the Western and Southern lines at Newmarket was as much as 27 minutes. That’s nearly half an hour to change lines. Why would you bother?
This is exactly why “plug in the density number” is probably not the best way of ‘predicting’ how good PT will do. Perth and Adelaide are two cities with similar Climate, similar population, similar densities. Adelaide has more train stations and slightly higher density. And yet it is Perth that has rail patronage that is a full five times that of Adelaide.
Perth and Brisbane have integrated ticketing (one ticket/card for all services) and we also have integrated fares (one price for all modes).
Ok, let’s go through Connell Wagner’s claims:
Why can’t Perth methods be adopted? Is this due to rail standards in NZ or a genuine physical constraint? I’ve tried to buy the engineering paper which has discussion on how to do this from Engineers Australia/Informit but have not had luck; The paper is here http://search.informit.com.au/documentSummary;dn=208925227741227;res=IELENG
Rail Achieving Growth: Conference Proceedings
Designing for Mass Transit Railways within Freeway Medians
Martinovich, Peter
This is probably due to gradient and curvature requirements. Light metro like Vancouver skytrain might be the solution here- it can handle sharper curves and steeper gradients apparently. Change the vehicle!
My understanding is that the Vancouver skytrain has shortish stations? Say 70m to 100m platform lengths can be accommodated by this kind of technology?
This might be true, but it is also true that with park and ride and feeder buses can expand the catchment area to something like 30 km! If the area around it is low density, that’s an argument for designing a rail service with tight integrated feeder systems, not giving up! Clearly the North Shore busway works now doesn’t it, so why wouldn’t a rail service with the trunk section replaced by rail and the feeder sections retained by bus work?
I was half expecting to see the “transfer penalty” argument pulled out in this one… Because Skytrain is driverless, you can also run very high frequency well into the night, and it has good peak hour capacity (30 000 pphd is more than Brisbane busway). So I think Light Metro should be considered.
Advanced Rapid Transit: Link is here http://www.bombardier.com/files/de/supporting_docs/Advanced_Rapid_Transit_en.pdf
Transfer penalty is a problem that Auckland must fix before anything like this can be considered. When there’s a direct cost to transferring, the rational behaviour is to avoid transfers.
Do you have to pay a new fee to transfer?
Ours is free for 2 hours. Other systems vary.
Yes. And everything is a transfer. Get off a bus, run into a shop and buy a chocolate bar, you have to pay to use the next bus that comes along from the same operator. Even if you bought a two-stage fare and you broke your journey before a stage boundary (don’t get me started on the static fare stages Auckland has for buses!) you still have to pay another full two-stage fare to complete the journey.
We have no all-modes, all-operator passes, and the closest we do have costs around $200/month. The all-modes day passes are about the same price as a five-stage return trip, so useful only to tourists.
Every change of vehicle is a transfer in Auckland, every transfer has a penalty. It’s killing the network efficiency.
“This is probably due to gradient and curvature requirements. Light metro like Vancouver skytrain might be the solution here- it can handle sharper curves and steeper gradients apparently. ”
Horizontal curvature is fine, grade is generally fine except for one section that would need a tunnel or deep cut, the main issue is vertical curvature. Basically the busway rolls up and down like a flattened out rollercoaster because they didn’t bother to straighten it out like the do the motorway and instead they did it on the cheap.
This really isn’t a major problem, all it requires is a bit of old fashioned cut and fill and a few extra retaining walls.
Going with light metro just to save a few million dollars of earthworks and retaining walls would be silly, really you’d want to straighten out the humps and dips with whatever mode.
Great analysis..but the impression I get from a Northern railway line is:
– Too expensive
– Feeder buses – one argument for a train is that we need less buses and trains are cheaper to run as only one driver & less maintenance, but you will need a fleet of feeder buses for the train to be successful, so there is still a huge operating cost for buses if you have the train.
– Buslane is working fine as is for population density on the Shore.
Your objection to this may be:
– Buses will clog CBD by 2041 – but if a underground rail loop is built for the CBD, then the buses only need to feed this loop. For example the Northern Busway could get to Victoria Park and drop passengers to a rail loop. Buses won’t be clogging the CBD.
– Buslane would reach capacity itself – then convert buslane to rail in 30-40 years time when we have the money and population density. Or perhaps impose a congestion tax so less cars which makes room for buses.
– Buslane in 2041 more expensive to operate than train – would need data to establish this, but the cost for the feeder bus fleet would be huge and needs to be taken into account.
To conclude:
– CBD – underground trains most definitely
– Outer suburbs – dedicated bus lanes feeding the CBD loop
One thought I’ve had was to feed between NBW services and a light rail service that runs through Viaduct Harbour, Queen Street, K’Rd, Ponsonby, Three Lamps, Victoria Park. Take the buses right off the inner city streets, add in a service that is generally popular overseas and is low-pollution (visual and physical), and you’ve got a big win for capacity and general design.
It’d also be a hell of a lot cheaper than building heavy rail to the Shore, no matter how it’s done.
Ricardo,
For one forget about population density. It really doesn’t come into the equation. Auckland is already the second most dense major city in Australaisa. It is far more dense that Brisbane, Perth and Melbourne which have extensive and well used rail systems and city loops, and only slightly less dense than Sydney.
All you need to know is that a quarter of a million people live on the North Shore and a quarter million is more than enough to support one rail line.
On feeder buses I think you are completely missing the point. The requirement for bus feeders is exactly the same with busway or rail, the only difference is with a busway you need a huge amount of extra buses to run the trunk route also.
To run a shore line every five minutes would require around fifteen trains plus around fifty feeder buses, to run the busway and it’s feeders to the same capacity would require over 250 buses. So with rail you have sixty-five vehicles and drivers underway at any one time, with busway you need two hundred and fifty vehicles and drivers.
Or think of it this way. With a busway you have one bus that collects people across a suburb for fifteen minutes or so, then it heads down the busway and on to the city streets to drop people off. It then heads back out largely empty, and gets back to where it started about an hour later to start again. So your one bus and driver can manage about two runs during the two-hour morning peak, or moves about 100 people at maximum.
If you have buses feeding rail, then your one bus simply just collects people across the suburb for fifteen minutes, drops them at a station and goes right back to collecting more people again and dropping them off again. So here your one bus and driver can manage eight runs during the two-hour peak, i.e. it moves about 800 people at maximum. So in this case you are getting eight times the efficiency by using bus to rail connections.
Oh and the busway is trending toward hitting capacity well before 30-40 years time.
The usual method of deploying buses appears to be like leaves on a tree, all feeding into stations. That’s fine for travelling longer distances along the trunk, but it is horrendous for more routine local(ish) journeys. The current construction of routes on the North Shore sees me able to walk from Birkdale to Mairangi Bay (7.8kms) in only 5 minutes more than it takes to ride (2 or 3) buses between the same two points.
The solution I’d think should be feasible would be to run feeder buses in arcs – even loops – that touch the trunk at two or more points….facilitating access both to the trunk and between local destinations for which the trunk is irrelevant. Also, running the service in both directions (clock-wise and anti-clockwise relative to the trunk) would make riding the service more efficient for all destinations.
I’ve noted that the Kaipatiki Rd Bridge has been there for a decade, but no Birkenhead transport services use it. They all hub to highbury…and then transfer to other desinations. A Birkdale or beach Haven resident may live only 20 mins by foot from Glenfield Mall…but must endure over an hour on *two* buses to get there via public transport in any configuration currently offered. I’m sure this is replicated all over the North Shore and it definitely puts me off using public transport.
I walk the 15.6km return journey to work each day instead of taking the bus. It takes about the same time and it costs me nothing. If the bus could do it in even half the time, I’d probably take it.
It’s not just the Shore, it’s all of Auckland. Because transfers are to be avoided at all costs, the route designs are horrendous. If transfers were costless, routes could be designed using more of a grid system such that even if you have to transfer it’s probably only once and the frequencies should be higher because each service is only operating over a fairly limited distance – as opposed to the current situation of some routes being 20, 30, 40km (some of the non-express services between Britomart and Papakura must be horribly long by distance travelled) long and thus requiring many buses to achieve even a moderately acceptable frequency.
…..and burning stupid amounts of diesel per passenger, inefficient, uneconomic, and all in the name of competition.
In some of these routes, say a so called express from Torbay, you effectively have a whole bus and a driver on staff to make one run inbound in the morning and one run outbound in the evening. The routes are so long by the time you get back out the peak is over and done with. So a half million dollar vehicle and a some $50k a year in staffing costs to get forty odd people to work on the weekdays. Not very efficent!
“So, for me, it’s heavy rail up the busway.”
I’m interested in the way that previous decisions shape less than ideal future choices. A motorway should ideally bypass densely populated areas while a rail line should service them. But in this case one is determining the other.
Similarly a freight rail line should bypass residential areas completely and service freight depots and ports. But both Wellington and Auckland are having to operate commuter services over a freight infrastructure.
I think one of the benefits of light rail is that it removes confusion between freight and commuter rail. You have more flexibility to run light rail away from both freight rail and motorway corridors and hopefully end up with a more attractive product.
@ NickR – you’ve given a well thoughout response – but to come back at you:
1) Population density – you say that the population density on the Shore is adeaquate for a rail line…OK, but not for the proposed price tag of $11.5 Billion..maybe it will be justified in 40 years time.
2) Feeder buses – I understand your logic about feeder buses being more efficient…but you could replicate that same scenario with a bus network…you could have feeder buses working their own suburbs, which then connect with high capacity (double length buses) for the trunk route and efficiency would be great…I agree that the operational cost will be higher for the extra drivers required, but when you look at an $11.5 Bill price tag for rail, then what’s a couple extra mill on bus driver salaries…drop in ocean..
To sum up: This bus network will then be in place until population density justifies the $11.5 Bill price tag, at which point it will be coverted to rail.
Except that $11.5b is the cost for an entirely tunnelled network with no existing rights-of-way and all the attendant costs. Building out the NBW into a rail corridor would cost a fraction of this because there’s already a right-of-way.
Yes improvements to feeder buses to the busway is something they should be looking at already for those same reasons. However even articulated buses are still going to have much the same inefficiencies and the same congestion on city streets. You’d still need nine or ten articulated buses to do the work of one train, and that is still two or three jumbo buses you need to send along Fanshawe or Cook St every minute.
Lets get something straight Ricardo, there is no $11.5 billion dollar price tag for Shore rail.
That is the cost of an entirely underground 18km metro style line from Newmarket to Albany, including a second CBD rail tunnel with four new CBD stations, a harbour tunnel and seven or eight underground stations on the Shore. Their costings are insane, they have priced each suburban station at a third of a billion bucks for example. This is double the cost of the proposed station underground in the middle of Albert St. In short it is a completely ridiculous over the top proposal cooked up by NZTA to discredit the idea of rail to the Shore, and their subterfuge appears to be working on you!
A far more realistic option would cost around $2 to $2.5 billion dollars for a harbour tunnel, a short CBD side connection to the existing network and conversion of the Busway to rail with an extension to Albany. This would give exactly the same capacity at a fifth the cost of their super underground system.
And forget about population density, it has no bearing on anything for Auckland. Population density would never justify $11.5 billion, it would never justify any amount. It really is irrelevant when you have all of the Shore within about ten minutes bus, car or bike trip of a station. Those 250,000 people is all the population catchment you need. Sure we can look at growing new dense centres around the stations, but it is no way necessary to do so.
I have to say, I am surprised at the cost of these underground stations. The cost should be reviewed by an independant group and have real-world reference costs. The estimates for Brisbane’s Cross River Rail stations in the Brisbane CBD came in at $100 million each. I struggle to see how a train station could cost more than that.
Even Brisbane’s Cross River rail was costed at $8 billion and that is an almost 10 km tunnel underneath the Brisbane River! So how can a rail line on the North Shore Cost more than that and be on the surface and in a pre-existing busway??? Huh?
If light metro or metro is used, the tunnels could also be steeper and shorter as well, reducing cost.
I don’t think it is a 100% substitute. Longer buses might have a capacity of 180 people and may be restricted in their running movements. So yes you can do this and it might be good as a stop-gap, but not a full solution. We already have tri-axle buses that can carry 100 people in Brisbane and also gigantic articulated buses.
It is not just about density. I can hold density constant but if I increase the catchment area from 800 meters to 35 kilometres by using park and ride and buses and cycling, then I have more people, even though the density is the same.
Mandurah line is a very poor comparison I think (tho I’m guilty of having made it myself previously): compare the actual distance you’re covering by this line compared to Perth’s southern suburbs railway. Compare how filled out the suburbs it’s passing through are. It would be a reasonably comparison if you decided to run the north shore line all the way to warkwarth, but I don’t think anyone suggests enabling that kind of sprawl do they?
I actually agree with the sentiment that metro style line is the best way to go here, the distances are short, you want to provide high capacity linking the main centres from Albany down through Takapuna, and into the city, with higher density living around the stations and particularly Takapuna. I don’t think it needs to be all underground. One thing to consider with the costs is the use of smaller stations, another thing is automation, for example Vancouver’s recent Canada Line which has small trains with small stations (~40m IIRC) but very frequent.
Now I still can’t understand why if you were to support a more metro style line that is largely underground, then why you wouldn’t go via Devonport and up the peninsula with stops that would be surrounded by existing suburbs perfect for infill and intensification. You’d get so much more on your investment in such a line than going the other way just for a stop somewhere in the middle of a giant motorway interchange at Onewa Rd, so that what, people can transfer from bus to railway to get into the CBD? That makes no sense to me, in that case just continue those busses across the bridge into the CBD. Run the busses in a pattern that weaves the metro line stops. Even the stop at Wynyard Quarter I don’t think would weigh that route in its favour in that case, that could be established in other ways.
Perth did it for $1.6 billion, 70 km, much of it in freeway median.
Since your rail is going down in the busway, the costs are comparable.
A good PT system will automatically attract density around stations. People will move and developers will cash in on that. Good PT first, urbanism later.
A metro style line, being separate, would not suffer from the decrease in frequency, and thus capacity, that a branching/forking rail line has. Vancouver skytrain can apparently reach speeds of 90km/hour if you really wanted to. Metros can be run on the surface, but also elevated (cheaper than underground) so you might use a bridge rather than a tunnel to get into the CBD if you wanted to. Fraser bridge I think it is called is the Vancouver one.