Next time someone tells you that transit has to be rail in order to affect real estate demand, send them this paper [paywalled] by Elin Charles-Edwards, Martin Bell and Jonathan Corcoran – a dramatic example of bus infrastructure profoundly transforming residential demand.
Our scene is the main campus of University of Queensland, which is located on a peninsula formed by a loop of the Brisbane River. It's in the southwest corner of this image. The area labeled "Brisbane" is the highrise downtown. Most everything in between — which is mostly on the south side of the river — is dense, redevelopable inner city fabric.
If you look closely you can see a single faint bridge connecting the University across the river. This is the Eleanor Schonell Bridge, which opened in 2006, and which is solely for pedestrians, cyclists, and buses. No private cars. It's one of the developed world's most effective of examples of a transit path that is vastly straighter than the motorist's options.
Prior to the opening of the bridge, University of Queensland had a problem much like that of Vancouver's University of British Columbia. Its peninsula setting helped it feel remote and serene (the rarefied air of academe and all that) but it was also brutally hard to get to, especially from places where students and lower-paid staff could afford to live. While there are some affordable areas west of the campus, most of the immediate campus area is far too affluent and low-density to house the university's students or the bulk of its workforce. So commutes to the campus were long and difficult.
Apart from the geography of income, the issue here was classic chokepoint geography, and that was the key to the transit opportunity. Brisbane's looping river, and its extreme shortage of bridges outside of downtown, slices the city into a series of hard-to-access peninsulas. Motorists are used to driving way out of direction to reach their destinations, and until recently, buses had to do the same thing. The only transit that could do what cars couldn't was the river ferry system, CityCat, and while this system is immensely successful, it is still a small share of the travel market because (a) so much of the population is not on the river and (b) the river is a circuitous travel path as well.
Charles-Edwards et al show how the bridge created an explosive expansion of access (where can I get to, soon?) for the campus by walking, cycling, and bus service. Walking:
And by bus (focus on the triangle in the centre of the image, which is the campus):
It's worth noticing why this bus bridge is so effective: It plugs right into the Brisbane Busway network, which looks like this. ("UQ Lakes" is the campus stop.)
Direct buses from campus run along most of these paths, and connect to many other frequent services covering the area south of the river, including a couple of useful frequent rail lines extending southeast from downtown. This is the biggest and highest-quality busway system in the developed world, in terms of the degree of protection from private car traffic along complete travel paths, including a tunnel under downtown. So the access opened up by this bridge was extraordinary. The busway is so fast and reliable that even commutes from northern Brisbane — on the same side of the river as the campus — were speeded up by the new bridge because they could remain in busway for the entire journey.
The effect on the location of students and staff, from 2003 until 2012 (six years after the bridge opened) looks like this.
The colour choices are unfortunate, so pay attention to the legend and focus on "St Lucia" (the campus) and the inner city areas just across the river from it. Remember, too, that this is a map of percentage change, so don't be distracted by big colors far away from the action, which represent noise (percentage changes on a tiny base). You can see that students and staff have shifted in big numbers to the inner city across the river from the campus, but also to southern and eastern suburbs each of the river, which are more affordable and still easily reached by buses from the campus. In the author's careful words, the bridge caused "a significant redistribution of staff and students across the metropolitan area." It also had the likely effect of reducing overall commute times by enabling people to live much closer to the campus, though the authors don't mention that.
Because the project gave buses so much of an advantage in accessing the campus, the mode share shifted dramatically, enabling the campus itself to grow without choking on cars:
Between 2002 and 2011, the population accessing the campus increased by 23 per cent, … all of which were absorbed by [non-car modes on] the bridge. There was an accompanying shift in the modal mix of trips away from cars to public transport. This was most marked among students, for whom less than one-quarter of trips were by car in 2011, down from two-fifths in 2002. Bus patronage increased among students from around a quarter of trips to more than half. Staff car usage declined from 70 per cent in 2002 to just over half in 2011, with buses, cycling and walking all increasing in popularity.
Eleanor Schonell Bridge is a powerful example of infrastructure that transforms a city's living patterns by transforming the isochrones of access. We can all think of trains and ferries that do this, but it's rare that buses are allowed to succeed in the same way. Once again, Brisbane has shown that it's not the transit technology that matters to people's location choices. It's where you can get to easily.
Brisbane has gone a great job of turning around the fortunes of public transit, and the busways to seem to be very popular.
I know Paul Mees was critical of the busways, as he said they just duplicated the suburban rail network, and that buses should have just fed into the rail network instead.
But the people seem to love the busways, and the commute time reductions are impressive in some of the corridors.
I found the bus ways a bit hit and miss. Miraculous for providing fast bus-only access to places like UQ, but prone to bus traffic jams at some places such as the main link out of the downtown/CBD area, eg:
If it helps visualise the UQ bridge, here’s a photo from the campus bus terminus:
Thats the thing with busways, and Ottawa has congestion problems. Although Ottawa’s issues are party due to never building a tunnel downtown.
From what I heard, Brisbane had to optimize bus service to reduce the number of bus trips on the CBD stretches, due to those bus congestion issues.
Brisbane is a bus success story all right, but it needs to be kept in mind that (contrary to received wisdom) this has not been achieved at substantially lower cost than an equivalent expansion of the rail network.
By and large, the cost per kilometre of building busways in Brisbane has come out similar to the cost per kilometre of building heavy rail lines in Perth. It demonstrates the fundamental fact that most of the cost of a high-performance transit system is in the provision of a dedicated grade-separated right of way, and this is largely independent of the mode actually employed. As an example, a branch line from the existing railway at Dutton Park across the Schonell Bridge to the university would have required just 1500 metres (one mile) of track and added at most 10% to the $226 million cost of providing the right-of-way for buses.
Some 15 years after Brisbane commenced its busway experiment, there’s a growing belief that the busway system isn’t sufficiently future-proof, the way an expansion of the parallel rail system might have been. Those famous one-bus-every-few-seconds convoys are now starting to look like just another traffic jam.
Tony, are you talking capex or opex or both? The latter is important and I note from annual reports that the gross cost per rail trip in Brisbane is circa $15 versus only $5 for the average bus trip. So even if CAPEX costs are the same for busway versus rail, OPEX for the latter *seem* higher.
Also, a branch line from Beenleigh line to UQ would not have been as effective as the busway. Reason being the busway provides access to areas to the south, east, and north much more directly than the rail network, which is much more circuitous in all directions.
So yes a rail connection to UQ could have been provided. But it would have also cost more to build and more to operate, while delivering less accessibility/patronage. Sounds like the busway connection was the better option!
As a regular patron on these buses, yes, the busway is often congested, and some of the bus stations don’t function too well in peak hour. But that – to me – doesn’t mean it’s a failure per se.
Significant investment has occurred to expand the network, which is great – though some of it could have been done much more cheaply and effectively simply by converting road lanes (but that’s another issue).
Possibly the main failures have been in not addressing central congestion areas.
Unfortunately, under engineered approaches to downtown are a hallmark of Brisbane transit design.
You find crippling congestion on both the rail and busway networks in the inner southern suburbs. In both cases the culprit is a poorly designed approach to downtown consisting of a standard size suburban station feeding via a narrow river crossing into a low capacity flat junction on the way to the main downtown station.
So yes, congestion on the busway makes the successful buses less of a success than they might have been. But if it had been built as rail, it would have likewise had it’s success limited by the same sort of congestion that exists for the same reasons. It’s not technology choice that is the story, rather Brisbane’s dedication to poor design on it’s inner suburban network infrastructure that is the story.
Tony Morton, are you joking? 10% extra from $226m for heavy rail! Not a chance. You clearly haven´t been through the tunnel and seen the alignment. Not to mention that there is no room on the Merivale Bridge for additional CBD trains. You could connect to the Beenleigh line to the south, but that means a double change for the majority.
I endorse Stu Dononvan´s comments.
Regarding many commenter’s comments on the CBD congestion, this is certainly an issue, particularly at the Melbourne St bus portal. Recently, I received a letter from the Director-General of Transport and Main Roads, responding to many issues I´ve raised. On this point, he said that it was a matter for Brisbane City Council. The Director-General is right, unsurprisingly. Specifically, they can easily increase capacity at this intersection by ejecting the West End bound buses into the general traffic lane (and Toowong bound turn right onto Grey St). Whether their inaction on this point is due to politics (would be unpopular with the pro-car lobby), incompetence or a more nefarious agenda, I can only speculate.
Not disagreeing with anything in the post, but extra comments for context:
– Brisbane’s busways, thought generally regarded as successful, have been extremely expensive, and there are some who say that it might have been better to put some of the money into improving the suburban train lines which already existed nearby. The chosen approach arose not from any sensible planning considerations, but rather from the longstanding lack of coordination and cooperation between the bus operator (Brisbane City Council) and the train operator (a vertically integrated state rail authority).
– The busways are a huge investment in the commute to the city centre. Brisbane’s public transport remains fairly dysfunctional for ‘anywhere, anytime’ travel.
– The Eleanor Schonnell bridge terminates in a turning circle on the west bank of the river. It would be EXTREMELY useful if some of these buses could be through routed to other places further west. The chosen design deliberately makes this impossible: a bus-bus ‘connection’ requires walking about 500 metres through the university grounds to pick up another bus on the west side.
I assume it was done this way because the promoters (or maybe the good folk of St Lucia) feared that a through road would come under pressure to be opened up to general traffic. If it were a light rail line, that problem would not have arisen.
Even allowing that Eleanor Schonnell buses can serve a range of eastern suburbs directly, it’s quite possible that overall network connectivity would have been better served by a light rail line from Park Road (nearby rail station on the east bank) via the university to Toowong (nearby rail station and activity centre to the west, roughly where the letters ‘UQ’ appear on the posted Translink bus network map).
I’m not claiming that light rail would have actually been practical in this case. Just an example of the political colour of some of the factors in bus-rail choices.
 St Lucia: the well-to-do neighbourhood immediately west of the university
Jack Horner, the lack of through routing came about due to the insistence of the UQ Senate which would not allow the bridge/bus station to be built if through routing would be possible due to the increase in traffic, even if bus only IIRC. That still doesn´t explain the awful location of the station on the far side of the playing fields. Could have been moved 200m west and saved 200m walking for 99% of users.
Exactly Simon Lovell, UQ Senate has planning authority within the St Lucia campus. When the government first proposed the Eleanor Schonnell Bridge, they needed planning permission from the UQ Senate for the western abutment. The UQ Senate initially said they would only give their approval for a bicycle only bridge, on the grounds that only a bicycle bridge gave adequate protection against transit being through routed through the campus.
Eventually a compromise was reached that allows buses on the condition that the station be a turn back design, quarantined from the road network and as close to river bank as possible.
Jack Horner, whether UQ might have allowed a turn back train or tram stop is arguable. But assuming they would have been happy for through routed transit under any circumstances is counter factual.
Take away through routing, and buses make the most sense. A one stop heavy rail spur that, due to inner city congestion, would require a transfer just to get to Roma St would be of limited success. Likewise trams.
Jack Horner. The dysfunction between the different government departments that run the trains and the buses may be well documented, but it isn’t particularly relevant to the choice of busways over rail. Mostly it isn’t relevant because the busways were built by the department that runs the railways (Queensland Transport) and not by the government department that runs the buses (Brisbane Transport).
As far as cost goes, yes the busways have been expensive to build individually. However you must understand, Brisbane grew around 18 radial light or heavy rail routes. In 1969, 12 routes were removed leaving big gaps in the transit network. The busways were born out of a planning study to find a way to plug these gaps.
They were chosen largely on the economics. Replacing the twelve lost light rail lines would have required a similar number of new rail lines, whereas the same job could be done with just three open busways and bus priority on Coronation Drive. Of course twelve new rail lines would have more capacity, but the density in Brisbane didn’t call for that much capacity. So, as a whole network, the busways were more economical.
Friend, this is an interesting take on a bus route, but even if Zeus himself came down and improved the transports and such in Brisbane Town, nothing, and I mean nothing, can change the fact that the fine CBD is built in an island enclosed on three points of the compass by a big ass muddy river and that the whole city has been built heap upon heap on ghetto ass suburbs in all directions with any lack of foresight and then them population ponzis and such mean that y’all are fighting hundreds of thousands to commute into the one location every morning and fighting to leave every evening and it makes me sad friends. This combined with the fact that the only business in town is the gubmnt means that Brisbane Town’s problems are only going to get worse. But lets keep them “prestigious” “inner” suburban suburbs as positional goods friends, y’all can’t let them young folks think that Brisbane Town aint the capital that is the Most Important One. I don’t even leave the fine CBD, that’s how bad the Brisbane Town ghetto ass roads and suburbs are friends, and no busway aint going to fix that up now. God Bless the fine folks of Brisbane Town, and God Bless Billy J. Jack the Duke.
Only game in town is gubmnt? Not true. Although it is true about the river hemming in CBD on 3 sides. CBD fringe is the main bugbear in Brisbane. Finally some reasonable rail through routing to South Bank has been implemented. Hopefully they fix up Milton soon. I´m much more optimistic about BNE than I once was.
So, if you read this data carefully, you see that the effect of the *pedestrian link* was *primary*, with the effect of the bus link being secondary. (Walking should always be first priority, for fairly obvious reasons.)
Here’s the thing: Brisbane has spent a fortune on the fanciest busways in the world. They’ve had decent results. Would they have gotten better results from trams? Of course they would have. It’s not that busways don’t work; they do, sort of. It’s that trams are cheaper to operate (more people per vehicle, electric power, longer-lasting vehicles), more popular (because they have better ride quality), and often cheaper to build as well (narrower ROW required).
“By and large, the cost per kilometre of building busways in Brisbane has come out similar to the cost per kilometre of building heavy rail lines in Perth.”
Brisbane decided that they would do well with buses fanning out without bus lanes in the suburbs, because they had low traffic there. As a result, the busways made economic sense.
This turned out to be wrong in at least one place, where the buses queue up at the end of the busway.
There are probably a few other places in the world where this model will work. But not very many, because they have to be relatively low-population (otherwise you need exclusive lanes in the suburbs too), extremely evenly sprawled through the suburbs (otherwise there’s no point in an ‘open system’), but big enough and congested enough to justify busways which cost more than rail downtown. This is an unlikely combination. It’s not even clear it applies to Brisbane.
“It’s not technology choice that is the story, rather Brisbane’s dedication to poor design on it’s inner suburban network infrastructure that is the story.”
Well, certainly, but constructed, expensive busways (as opposed to take-lanes-away-from-cars bus lanes) are inherently poor design. The busway is simply an example of dedication to poor design, one among many.
Jarrett’s a technology fetishist who prefers buses, and it’s become bizarrely obvious over the years, despite his protestations to the contrary. This is a good reason not to hire him, unfortunately.
Nathanael wrote: “Would they have gotten better results from trams? Of course they would have.” Not one shred of evidence for this. I admire the ability to *know* what would happen in a “what if” scenario.
Nathaniel. You claim that dedicated busways fail because they don’t have sufficient capacity to work effectively, such as the at grade bus lane section of the SE busway on Melbourne street.
You then claim that a better bus based solution to capacity problems like those afflicting the at grade bus lanes on Melbourne Street is … at grade bus lanes (presumably like the ones on Melbourne St).
If you are going to point to a single poorly designed stretch of busway as representative of the whole concept, then I shall point to the short stretch through Mater Hill station which carries over 18,000 ppdh and use that to justify a claim that buses clearly are capable of moving 50% more people than the largest tram system that would fit within the station footprint. It’s only fair after all.
Mater Hill does not carry 18,000 pphd. The only part of the system that gets near this is the very, very, very short stretch between the Woolloongabba junction and the Captain Cook bridge on and off ramps, which has no stations and is little more than a massive single point of failure. Mater Hill gets around 10,000 at maximum based on the current services delivered, and this relies upon most of the buses being no longer than a 12.5 metre rigid (60-70 passengers max) because the platforms are too short to accommodate 2 high capacity buses at once. That is broadly comparable to a light rail system operating with Gold Coast-length vehicles at 2 minute headways.