Friday, September 30, 2011

Good, Bad and Interesting things about Adelaide Transport

Presenting a paper at this year's Australian Transport Research Forum in Adelaide has provide a chance to sample that city's public transport. Here's four good, four bad and four interesting points based on observation.

The good

Go Zones. Frequent service corridors covering most inner suburbs out to about 10km from the CBD. They are extensively advertised at stops, on timetables and at the Metro Shop.

Airport accesss on regular services. J1 and J2 provide a 15 minute service 7 days a week. Service spans are very wide, with service starting before 5am even on a Sunday morning. The profile of the service is quite high - airport staff recognise the numbers and the information desk is well stocked with timetables.

Rail electrification. Project includes several new and rebuilt stations, sighted on the Noralunga line.

Glenelg tram. New extension is well patronised. It also serves major trip generators including a university, convention centre and hospital under construction.


Pedestrian crossings. Imagine a journey where after a couple of minutes travel you stopped, were paused 2 minutes, could lurch forward a few hundred metres and stopped again. This is walking in Adelaide. Long traffic light cycles at CBD intersections reduce overall walking speeds to a crawl. In the suburbs islands and seperate signals (for each direction) at divided roads further slow transfer between train and bus. The Melbourne equivalent would be if every intersection had traffic light cycles like King Street.

Infrequent trains. Unlike Melbourne or Perth, where trains form the most frequent 'spine' of the network, train frequencies are often 30 to 60 minutes, making recourse to a timetable essential.

Low bus network legibility. It is difficult for the visitor to see the logic of the bus network. If you board a bus in a CBD street you cannot be assured it will continue straight along it. There is a large number of route numbers, with various letter and number prefixes and suffixes. Buses are significantly less legible than trams in Melbourne, but there are no inherent reasons for this to be the case.

Few maps on the network. Compounding limited legibility is that while many bus stops have times, few have maps of either the route or network. The only place where there's a city-wide network map appears to be inside the Metro Shop. Go Zone network maps are similarly available on the web but not at the point of need on the system. Maps of individual routes don't seem to be nearly as common as (say) Melbourne.


Ticket purchase on trains. Instead of at machines at stations.

Can see to the front on trains. Most systems' trains only allow passengers to see out the sides of a train. With at least some of Adelaides you can also see out the front. This gives a quite different view of the network.

Single zone tickets. The liability is a high minimum fare, though there is a cheaper short-distance ticket. The advantage is simplicity. The ratio between single and daily ticket is not dissimilar to Melbourne, making a daily tickets a good choice.

Stops are numbered. The acid test of a public transport system's legibility is whether people can find themselves to a destination at night. Large numbers on stops are viewable from the bus, so can help if trying to ascertain where you are (printed timetables refer to these numbers against timepoints). On the flip side timetables are not stop specific - they are instead full timetables where the passenger must estimate arrival times for themselves.

Saturday, September 24, 2011

Why even culs-de-sac with walkways are bad

There's been some debate on culs-de sacs. We all know they're terrible for pedestrian permeability.

But is providing a walkway at the end of them a panacea? Some have suggested that this provides the best of both worlds - safety for children due to the removal of through car traffic - and permeability for pedestrians.

While better than culs-de-sac without end walkways, they present some problems. Overall I think they're inferior to a grid streets (with narrowing and traffic calming where necessary).

I give five reasons why in my response to the previous link:

1. Opportunities for graffiti/vandalism. Cul de sacs create blank side fences. These are a canvas to vandals. Urban design that minimises these is good (for the same reason there should always be verges, roads and then houses facing railways – never back fences).

2. Poor passive surveillance compared to a continuous street with house frontages. Increase opportunities for vandalism and assaults against pedestrians. Criminals think they can ‘get away’ with more if there are no sightlines.

3. Low information and legibility. Walkways are less prominent in the street directory and only locals may know about them (compared to continuous streets). Trip planner type mobile apps that only consider streets that cars run on might not have detailed pedestrian access way data.

4. Impermanence. It’s not only in posh areas that cul-de-sacs can be closed. To take a random example, when it was built c1979, the end of Wimmera Court (Werribee) was open.

Now it’s closed and part of private property. Since then a shopping centre opened. Had the end remained open it would only be 10 – 12 min walk from it. Now it’s nearer to 15 min, with the perceived time longer due to less directness. I contend that a legible 10 min walk vs a less legible 15 min walk is a huge difference in walking’s attractiveness and thus its share.

Police and residents may push for closure due to apparent crime problems (see 1 above). But especially if there’s a wider pedestrian access issue that affects others from outside the street, then such calls should be resisted due to its effect on the pedestrian network. But it would have been better not to build culs de sac in the first place.

And I don’t think we should just concentrate on access between schools/shops and houses, even though this carries the higher volume – we should also consider anywhere to anywhere trips – eg kids visiting friends houses.

5. I don’t know if it’s my imagination, but in poorer areas the most derelict houses with the worst gardens, the most number of cars on blocks and sheets for curtains are gathered at the ends of courts. If that’s not factually the case it looks worse around the bowl of a court. Check out Studley Court, Laverton on google. Also front fences on houses at the ends of culs-de-sac can be casualties of wayward cars as well. Though not a believer I do think Feng shui has some good design principles, and I that a house on a cul-de-sac bowl draws the shortest straw.

Monday, September 05, 2011

Should transit follow people - or people follow transit?

While not the exact subject of the post, a recent Human Transit post contains a general quote that's worth reflecting on.

This, to me, fits into a much larger agenda of insisting that everyone who makes a location choice -- especially about where to live -- should be required to acknowledge the transit impacts of that choice. Today it's still common to encounter the other sequence, in which someone (a) signs a lease or deed of sale or development agreement and then (b) yells at the transit agency because the service isn't what they desire.

This is pertinent in any city where transport service quality varies.

Here in Melbourne there are roughly three service tiers correlating with inner (0-10km), middle (10 - 30) and outer (30 - 50km) suburbs. The first is dominated by trams, normally on a 10 to 20 minute frequency, although a few ex-tram bus routes also qualify. The second has trains and/or higher service buses (15 to 30 minute frequency). Then there's the outer suburbs with buses and sometimes trains (40 to 60 minute frequency). To this may be added an exurban and near regional zone, whose service varies from nil to hourly.

The above is a broad generalisation and exceptions apply. There exist inner and middle suburbs (Kensington Banks, parts of Port Melbourne, northern Reservoir, parts of Sunshine North) with limited span, frequency and coverage. In contrast some semi-rural areas (Warrandyte, Yarrambat) have almost inner-type service levels. And others (Woori Yallock, Cockatoo) have buses running as late as in a regular suburb.

These are not academic discussions as service planning decisions have human impacts. These are disproportionate if a service (even if temporary) is introduced, relied on by a few, and then discontinued.

Take Kinglake, for example. It's a semi-rural lifetyle dormitory area north of Melbourne. People move there to 'get away from it all'. They may be semi-retired or drive to work in Melbourne's northern suburbs or the CBD itself.

It had one suburban bus each weekday to the major suburban centre of Greensborough. The area was devastated by bushfires and the government introduced a frequent 7-day service to assist with relief recovery. Average patronage was not high and the added services were withdrawn. A residents' campaign called for services to be restored. The government responded, introducing a limited service, to start next week on a trial basis.

People in Kinglake (like those in Mornington Peninsula, Pearcedale, Warneet, Kangaroo Ground, Mt Martha or Eynesbury) pay taxes and expect government services in return (including public transport). Even an affluent car-owning population in these parts may request public transport on behalf of others, eg for their teenage children with part-time jobs or elderly parents.

Unfortunately dead-end routes to semi-rural areas are unlikely to be well patronised. They require high subsidies per passenger. If the aim is the 'greatest good for the greatest number' better uses for the resources are likely to be found elsewhere. If we temper this approach to acknowlege the 'tyranny of the majority' and provide a minimal 'safety net' service to sparse areas, the extremes of under-provision are avoided but at some opportunity cost (most notably in services foregone to lower income outer suburbs or to provide main road frequent service corridors).

Jarrett's quotation suggests we should have little sympathy for those who move to a poorly served area and complain about the service. One might also add the irony of affluent city folk seeking to 'get away from it all' in an exurban hideaway but then demanding city-type services (largely paid for by others).

Moving house is a major decision and it's not unfair to expect people research where they're moving to beforehand. Such expectation of personal responsibility should rise with incomes; those who earn more by definition have more housing location choices. Families, in particular, are used to making these types of decisions, especially in relation to childrens' schooling. Ditto for some older people, though a regular suburban house has more nearby services to 'age in place' than a high-maintenance semi-rural acreage, so the latter may require an earlier move out.

I should add an important qualification. Jarrett sees a world where bus service quality is based on objective criteria like pedestrian catchments, population density or a corridor's importance. If we planned like this services are likely to be more secure since there is a reasonable alignment between service provision and demand. And cases of gross over and under servicing (neither of which are sustainable long-term) would be fewer.

However in practice we know that service levels and network need do not necessarily match. I mentioned Warrandyte and Yan Yean, both of which receive more service than density or demand alone may justify. In Yan Yean's case though, the service is only as good as it is because it is on a major cross-suburban orbital route. Provided through passenger numbers were high this may be justified, with Yan Yean being a lucky but incidental beneficiary.

Should we be shifting some responsibility for public transport on the individual? Ie should people without cars live in a semi-rural area and expect urban-type public transport? Or should there be some obligation to consider transport in location choices?

I lean towards the latter, with two major reservations.

The first is that service levels are transparently determined and fairly reflect likely demand or need. That a route runs frequently only because it was an old tram, or that it does a special extension to a long-closed hospital would not be good service planning, for example.

Given the public subsidies involved, and the allied public interest, it could be useful to weigh up the merits of more widely distributing patronage data versus holding tight. Release would likely strengthen the hand of the service planner when pressing 'greater good' cases for service reform and educate the community about the trade-offs and opportunity costs involved. Such information would also be a useful antidote against those who may lobby to retain an indirect route deviation but never used it when it ran.

Secondly, there may be a need to provide coverage to some areas to 'complete the network' or 'fulfil social needs', even though these services may not be highly patronised. However their potential opportunity costs should be acknowledged by planners and shared with the public to better inform the decision made and increase acceptance of it.

For instance running an indirect route via a street with retirement homes in a low income area may be accepted (though ideally planning policies would only approve such developments near actual or likely bus routes). Whereas the strict '90% within 400 metres' coverage requirement could be relaxed in less needy or less dense areas (eg Brighton and Templestowe), especially where alternatives like trains and high-service buses are available to most within 800 metres.