Saturday, September 27, 2008

The Pedshed Series

or how to get the patronage gains of 10 extra stations without building them

Most people with a choice will only walk a certain distance to public transport. The generally accepted maximum is 800 metres/10 minutes to 'high quality' public transport (eg railways or busways) or 400 metres/5 minutes to 'regular' public transport (eg local bus routes).

The area surrounding a station within 800 metres or ten minutes walk is within its service area or pedshed. Since people are more concerned about time rather than distance, the service area of a station can be increased if overall origin to platform walking speed is increased.

Because walking speeds are a matter for personal discretion, the only way to do this is to eliminate all obstacles between station platforms, origins and destinations. The most significant of these obstacles are those where the passenger is forced to wait or is moving much slower than regular walking speed. Examples of obstacles may include unresponsive pedestrian crossings (or none at all), indirect detours, inconvenient entrances/exits and bottlenecks near ticket barriers.

The diagram below shows how a station's service area can be increased by improving access speed.

Because the increase in area is proportional to the square of the radius, a even small change could easily increase a station's potential catchment population by 10 or 20 percent. If pedshed enhancing measures were applied to all stations in a 200-station network, it's not unreasonable to expect patronage gains equivalent to building several new stations.

Let us get some more detailed figures on this.

800 metres in 10 minutes assumes a walking speed of 4.8 km/h.

Suppose we were able to speed station access times by 60 seconds. Instead of passengers being able to walk 800 metres in 10 minutes, they can now walk 880 metres in 10 minutes.

The pedshed areas are as follows:

800 metre pedshed: 3.14*(800*800) or 2 009 600 m2

880 metre pedshed: 3.14*(880*880) or 2 431 616 m2

This is an increase in service area of 21%. Patronage increase might not be quite this high (those 10 min away might be less likely than those 5 min away to use it) but even if half or three quarters the service area increase it is still substantial.

It works in reverse as well. Removing a subway or zebra crossing in favour of a signalised pedestrian crossing (for example) reduces access speeds and therefore a station's pedshed. In the example above, a minute's delay reduces the pedshed to 720 metres and coverage area to 1 627 776 m2 or a 19% reduction.

The other effect with signalised crossings is that waiting times vary, meaning greater variability in origin to platform times. Walking time via subway or zebra crossing is constant and predictable but becomes variable when they must encounter pedestrian signals (especially those on 90 second cycles). Assuming two crossings, worst case variabilty can be 3 minutes, which is a significant proportion of the ten minutes access time allowed. The effect is that passengers must build 'fat' into their schedule (by leaving early), further reducing origin to platform travel speeds.

A program to widen station pedsheds would require a range of measures of varying cost. These vary from subway or elevated walkway construction (most expensive) to altering pedestrian cycles, removing roundabouts, installing median strips or painting zebra crossings. However such measures are desirable if public transport is to operate at peak performance, delivering the lowest possible (and most predictable) door to door travel times.

Sunday, September 21, 2008

Welcome to Transport Textbook

I'm pleased to announce that melbourneontransit has joined Phin's Transport Textbook. This is like a group blog but the actual posts and comments remain at our own URLs. Think of it as an index that gives quick access to the latest from Phin, Riccardo and myself.

Saturday, September 20, 2008

How to get concession public transport and not be poor

Concession public transport is only for low-income people, such as age pensioners, disability pensioners, students and the unemployed. Right? Wrong!

This post will examine two cases where one can be quite well off (a millionaire household, even) yet get concession or cheaper public transport. It raises some welfare policy issues of who gets what, whether the greatest benefits go to the most needy, or if it's more about votes.

Chapter Six of the Victorian Fares and Ticketing Manual explains who gets concession and free travel.

While many groups get free or concession travel, in this post we will talk about only two; low income earners and seniors.

Low income earners

This group of passengers do not need to be on any welfare benefit, but they do need a Low Income Health Care Card, issued by Centrelink.

Maximum qulifying weekly incomes for various household types are shown here, but for our purposes an amount of $23192 for singles and $38636 will suffice. Income can go up a bit and you will not lose the card, but you won't be able to renew it unless it is below the qualifying amount mentioned above.

The income limits above are not particularly high. There would be many people working part-time (and even some full-timers) who could qualify. Since commuting is a major added expense when one goes to work, the effect of this benefit as it applies to public transport is to assist the 'working poor' and provide an incentive to get off welfare.

If income rises further (eg extra hours at work) a poverty trap effect is created as the concession entitlement cuts out and the passenger must pay full fare. This difference is about $23 per week, assuming Zone 1+2 travel. This is not unique to public transport, and also applies to any heavily targeted welfare benefit with a steep taper rate. The alternative in this case is lessening the benefit withdrawal taper rate, but the problem with this is that welfare becomes both dearer for the taxpayer and less targeted at the needy.

I see nothing wrong with low income working people getting concession travel on public transport, and indeed there are social benefits, especially if it improves work incentives.

However it gets interesting if the cardholder is not working, for instance living off investment income.

Since the Centrelink low income health care card has no assets test, only an income test, you can have quite a lot of assets and still qualify for one (and thus concession travel). How much? Well it depends on the deeming rates.

Working back from the income limits above the result is up to $386533 of assets for singles and up to $643933 for couples. This is a rough (conservative) calculation based on a 6% deeming rate; where 4% applies for part of the capital the figures will be higher.

On their own these are quite respectable asset levels to be deriving government benefits (or concessions). But that's not all. Since the family home is exempt we can add that as well. Assuming ownership of an average home (worth $400000), you can be worth nearly $800 000 as a single or over $1.2 million as a couple and still get transport concession benefits!. Or to put it in another way, you can be a millionare couple who owns their own home and gets $38000 pa and still get concession tickets since Centrelink deems you 'low income'!

Seniors Card

Sick of slumming it on just $38k pa so you can get concession fares? Well you've got a treat coming as soon as you turn 60! Provided you watch your working hours (because I doubt others will), getting a Seniors card is your key to a wonderland of benefits unavailable to working people.

Younger Centrelink clients report various letters, compliance checks and audits to verify you're still entitled to benefits (which is as it should be to maintain the welfare system's integrity). But if you've got a seniors card, about the only mail you'll get are free country travel vouchers. Plus you can rely on the local rag to convey news of seniors festivals (more free travel!) and other goodies.

Admittedly the above has been a bit flippant but the reason will become apparent later. But beforehand, let's distinguish between seniors and (age) pensioners. All age pensioners are seniors, but not all seniors are age pensioners. The single age pension at around $280 per week (single) isn't exactly a life of caviar. However at least they do get various Centrelink benefits, and the assets and income tests ensure there is a sound needs basis for them.

Seniors Card is a quite different animal in no way related to various needs-based Centrelink pension and health care cards.

Over 60? Victorian resident? Well that's two out of three criteria met. The other one is a rather curious work hours test. If you're delivering junk mail around the streets part-time you're in. If you've got your own surgery so can control your work hours to 34 per week you're in. Ditto if you're a BHP board member. But if you're a PAYE earner working 38 hours per week you fail. Eligibility boils down to the control you have over your time, and self-funded investors or business proprieters have more of that than full-time (and taxpaying) employees. Plus, unlike Centrelink cards, no assets or means tests apply.

Since its criteria is so rubbery, you'd think that Seniors Card would attract less generous conditions than than the concession fares enjoyed by holders of Centrelink cards (which have needs-based eligibilty requirements). Again you'd be wrong, at least for Melbourne.

Compare the daily ticket fares between a Zone 1+2 concession daily ($5.30) and the Seniors Daily ($3.30). The non-means tested Seniors Daily is nearly 40% cheaper than the means-tested regular concession tickets. This indicates that if the Seniors Daily ticket has a sound reason for existence it cannot be on fairness, equity or welfare grounds.

However my journey to find a rationale for the sub-concession Seniors Daily did not return empty-handed. This came about because a lot of senior travel is 'discretionary' or 'impulse'. And elasticity for impulse, recreational and other concession fare trips is higher than (say) another person's journey to work.

On all modes interpeak weekday service frequencies are higher than during the weekend or at night. Working people are working and youth are at school during this time. So that leaves the unemployed and seniors as the main potential interpeak passenger base. While services may not be crowded, interpeak patronage is still important to a successful public transport system.

And as the buses are already running, it seems to be in everyone's interests to stack them with sub-concession passengers rather than have them empty. The same seniors who so powefully resist a ticketing change can also resist a bus service cut, so it may be that it is in the interests of operators and public transport generally that seniors form a political constituency for a route's survival.

Since public transport relies on public funding it is inherently political and susceptible to government decisions with regard to services and service levels. A bus route full of seniors is less likely to be withdrawn than one that runs empty, so from the point of view of self-preservation it seems better to carry seniors cheaply, even if the equity is terrible vis a vis concession passengers.

This leads us to a conflict; the policy rationale for the Seniors Daily ticket is poor, but it has great value in getting 'bums on seats', especially during the interpeak period. Outright abolition of the Seniors Daily (while the best on equity grounds) could cost patronage and, given recent events, is politically impossible. Another option could be to maintain the Seniors Daily but confine its use to off-peak travel. This might shift a small amount of discretionary peak senior travel to the off-peaks, so slightly relieving congestion and increasing overall patronage and revenue. Plus equity is slighly improved vis a vis concession passengers as the latter's relative penalty is less ($3.30 vs $4.75 instead of $3.30 vs $5.30). Politically, event this may not be worth the trouble, despite its policy soundness.


I have illustrated two ways in which wealthy households can get concession (or cheaper) public transport.

The first is through the Centrelink low income health care card. Though some millionaire households (living on interest or rents) also qualify overall it is generally in the public interest that such cards continue to be good for concession travel because the main benficiaries would be the 'working poor'.

The second is through the Seniors Card, which can sometimes offer cheaper travel than available to concession holders. The Seniors Card (and thus the allied Seniors Daily ticket) has slack eligibility criteria, fails all equity tests and appears to be more about buying votes. The Seniors Daily however successfully promotes off-peak use and the retention of this ticket for off-peak travel would slighly improve equity while still promoting patronage.

Monday, September 15, 2008

The best of Melbourne public transport

A few of the 'bests' in Melbourne public transport services.

Best span

Skybus runs 24 hours. Runner-up is Route 896, the Cranbourne TrainLink, which is somewhere near 20 hours.

Most frequent daytime service

There's several tram routes that run better than every 10 minutes, with 19, 59, 86 and 96 rating highly. 19 takes the cake for its frequent service seven days per week and except Sunday mornings would win 'best weekends' as well.

For buses the new 401 from North Melbourne Station runs every 6 minutes off-peak weekdays. The other bus route with a 10 minute or better frequency is 246 up Punt Rd. Ditto for the Knox Transit Link portion of Route 732

Most frequent weekday peak service

401 again for its 3 minute peak service. Belgrave/Lilydale line trains manage a similar service level, at least to Blackburn. Second place amongst the buses is held, not by a SmartBus, but by the low-profile but well-used 465 between Keilor Park and Essendon with a 7-8 minute average frequency.

Best weekend service

Pretty much any tram route will win here with services at least every 15 minutes. This includes Route 82, our only tram route with a better service on weekends than during the week. This is followed by trains with a 20 minute frequency at most stations.

As for buses, some of the MBL 200-series routes do well (every 15-20 minutes), followed by Knox Transit Link 732, South Morang TrainLink 571 and the 700 SmartBus (on Saturdays).

Best night service

Skybus without a doubt, with the new timetable extending 15 minute service until late seven days per week. The only other bus to run a 15 minute night service is 220 between Sunshine and Gardenvale on weeknights. On weekends service is comparable (though earlier finishing) to trams, with a 20 and 30 minute Saturday and Sunday evening frequency respectively. Honourable mention goes to the new SmartBus Route 901 between Ringwood and Frankston, with weekday services every 15 minutes passing Knox until after 10pm.

Most trams run a flat 20 minute service until about midnight, with later services on Fridays and Saturdays and 30 minute service on Sunday evenings. Night trains on most of the network are a flat 30 minutes except for the ex-Hillside lines where they're every 40 minutes on Sundays. Caulfield is the furthest suburban station which receives a 15 minute night service until last train seven nights per week.

Best transport corridor

Swanston Street/St Kilda Road wins due to its length, the number of routes and combined frequency. Others to rank highly include Oakleigh to Chadstone, Doncaster to Box Hill, Southland to Cheltenham and Footscray to Highpoint and Clayton to Monash University, probably in that order.

Saturday, September 06, 2008

Connex Customer Information Centre opens

Yesterday was the first day of the Connex Customer Information Centre at Flinders Street Station. The centre is open seven days a week and is on the main concourse, inside the fare-paid area.

The opening presents a good opportunity to discuss the sort of customer service passengers need and how this is best provided.

Passengers are best divided into regular commuters (who make frequent routine trips) and occasional travellers (eg tourists and those who don't normally use public transport). The information needs of these groups vary, as per the table below.

Basically occasional passengers need detailed individual advice, while regular passengers just want their ticket or timetable with a minimum of fuss. At very busy stations it may be desirable to seperate the routine transaction from the trip assistance role so regular commuters are not delayed. The arrangements at Flinders Street appear to do this in that there are ticket windows for sales and the new information centre for more detailed queries.

Just like in a shop where you can ask questions about a product before buying, passengers need to be able to ask about a service without needing to buy a ticket first. This means that the main enquiry desk needs to be immediately outside the fare paid area, as all staffed suburban Melbourne stations do with their booking offices and Transperth does with its InfoCentre at Perth Station. Doing this makes the information booth accessible to those likely to need it (see table above).

As well as helping passengers board the right train at Flinders Street, there are other information needs that a full information centre can be expected to meet. This includes advice for the onward journey; passengers may alight at an unattended station and may need bus information for the next stage of their trip. Planning here may determine whether passengers board the next train or get a later one for a better bus connection.

Because no one (except gunzels) ends their trip at a railway station, Connex passengers alighting at Flinders Street need to be looked after for the next stage of their journey. This might include directions to a particular address or details tram and bus services. Best practice (as done in Perth) is to have all timetables available, but as a minimum in Melbourne all tram/bus timetables for the CBD, major trip generators and major suburban routes need to be carried.

Disruption information is largely electronically delivered through passenger information displays and PA announcements. An information desk may have a role, although sheer passenger numbers may not necessarily allow individual attention. The effectiveness of this depends on the extent to which information is conveyed to the staff and their awareness of possible alternative tram and bus services (both rail substitute and regular).

The concept of expanding customer information at major stations is to be commended. However placing the main information booth so it can be accessed off the street and providing tram/bus information are both essential for the idea to reach its full potential.

Wednesday, September 03, 2008

The TOD Top Ten

Despite concerns over fuel prices, resource scarcity, population growth and housing affordability, there is still large amounts of land near shops and railway stations that is not utilised to its 'highest and best' use.

While high-density apartment blocks in outer suburbs are beyond the 'human scale' and are not supported by market demand, there is much to be said for medium densities such as small-scale low-rise apartments, townhouses, villas and semi-detached houses.

Of these, the most promising is the villa. They were built in large numbers in the 1960s and 1970s certain suburbs such as Reservoir, Carnegie and Chelsea. Typically part of a small group (3 to 8) they provide adequate privacy and space for smaller househould, including a private yard, carport or garage. Because villas are typically single-storey, their construction and maintenance costs (ie body corporate) are lower than two storey townhouses or even multi-storey apartments. Despite this the space, amenity and ease of maintenance offered is intermediate between an apartment and a house and thus likely to be of wide appeal.

Where in Melbourne could such transit-oriented residential developments be located?

While people discuss about what they need in a suburb, the three big things are good transport, shops and schools. Villas are less likely to contain children than seperate houses so the first two are relatively more important than the third. For 'good transport' we can read train, and for shops, we can read 'large supermarket plus some specialty shops'.

We are lucky to have a large number of such suburban centres near railway stations in Melbourne. Unless you've been to Perth (where railway stations can have no nearby services except for bus stops) this benefit is not sufficiently appreciated. The sort of centres I'm talking about are generally Melbourne 2030 major activities centres and include places like Heidelberg, Mentone, Ferntree Gully, Glenroy, Carnegie, Noble Park or Chelsea.

However only a minority of the suburban population is within walking distance of a station and a smaller minority is near a railway station that has a reasonable retail and commercial area such as those listed above.

Increasing the proportion of the population near such well-serviced centres is the tenet of the Melbourne 2030 plan. In the public mind this was associated with higher densities in prime suburbs and frequently met with NIMBY resistance.

A look at any street directory indicates there are a number of areas that could attract higher housing densities but are not normally considered transit-oriented centres. My criteria for such potential TODs is that it have a railway station and supermarket within about 1 kilometre of one another. The area in between is thus walkable from both and could lend itself to more transit and pedestrian oriented housing. Minimum parking space regulations would be abolished in such areas due to the likely higher use of other modes. And this in turn could reduce costs for businesses and developers, providing a possible incentive to invest.

So without further discussion, here's my list of the top ten potential transit-oriented development areas:

1. Lalor. Some areas are within walking distance of the Lalor Shops and either Thomastown or Lalor stations.

2. Fawkner. Includes railway station and shops on Jukes Rd. Sydney Rd presents some barrier for pedestrians but some houses are within walking distance of both.

3. Keilor Plains/Centro Keilor. A small area is within walking distance of both. Main roads again present barriers for pedestrians.

4. Watsonia/Diamond Valley Shopping Centre. A busy road nearby but abuts the generally affluent and green north-eastern suburbs.

5. Narre Warren/Fountain Gate. Has the Southland problem - the shopping centre is just beyond a reasonable pedshed of the station. However some areas half way in between are in the pedsheds of both.

6. Berwick/High St Berwick. Main constraint is limited land supply in area in between as much is used for school, retirement and medical uses.

7. Merinda Park/Thompson Parkway SC. A small section of Endeavour Drive is walking distance to both.

8. Seaford/Safeway Seaford. A small area is within walking distance of both but again supply is limited.

9. Cheltenham. Only noble Frank Fisher-style public transport martyrs would say that Southland is within the pedshed of Cheltenham Station. However some land is within the 800m pedsheds of both. Much is currently taken up in car-oriented commercial uses.

10. Williamstown. Major shopping strip within walking distance of three stations. Already high-rise (housing commission) tower in area. Area considered historic, so future development would need to be sensitive to this.

Laverton just missed the list. While parts of the suburb are walkable from Central Square Shopping Centre (Altona Meadows) the freeway presents a major barrier and makes foot access from one or the other to nearby houses unattractive. Cranbourne is another possibility though the station and shopping centre may just be too far apart. People in medium and high density sacrifice personal space for access/amenity, and standards regarding the latter need to be higher than for houses.

Numerous places where shops are declining or stagnant present development opportunities. Examples include Moorabbin (once conceived of as a major Box-Hill sized centre) and Glenhuntly. Many smaller centres are restricted by the lack of a large supermarket; these include Murrumbeena, Edithvale, Oak Park and Blackburn. Kmart Campbellfield, though car-oriented, would present some opportunities with a station; and of course Southland is a no-brainer. High through traffic can sometimes reduce local centre amenity (Moorabbin and Ormond), but low through-traffic can also lead to decline (or may be an expression of it) eg Laverton and Patterson.

Other TOD opportunities include Westall Station around which there is significant new housing being built. This is best described as commuter-oriented development; unlike a true TOD there is no substantial retail. Bus services are limited and residents are almost certain to drive to get even basic food items.

Williams Landing (near Laverton) is another to consider. It illustrates that 'ending (and building elsewhere) is preferred to mending' as Laverton shops sit vacant (Zone 1 notwithstanding). If infrastructure is timed right it could be a true TOD, rather than a TAD (Transport After Development), as ocurred with Roxburgh Park and to some extent Sydenham/Watergardens.

The most outstanding TOD opportunity, however, is Caulfield. The transit provision is good, with train frequency never worse than 15 minutes until midnight seven days a week. Caulfield is a desirable suburb with a potential NIMBY element. However this is mitigated by an underused racecourse that has made itself unpopular with residents by restricting public access to its parkland. Evicting the racecourse (the land is publicly owned) and replacing it with 50% TOD (station end) and 50% parkland would be a fair compromise that would benefit existing locals and new residents alike.

To summarise, some readers will be intrigued that most of the areas listed above are regarded as 'cheap' suburbs well beyond the 'latte belt'. While Williamstown and Cheltenham are exceptions, most of the rest (eg Lalor and Fawkner) have little natural beauty or recreation and are a fair distance from Melbourne CBD (Zone 2) and so-called 'good schools'.

Reflected in house prices is that most of the above areas aren't exactly high-demand suburbs. However that won't stop people moving into them if they're the best of the affordable choices and offer all 'must have' services. Rail is the clincher; rail suburbs like Lalor or Keilor Plains are better commutes to the CBD than desirable but bus-only suburbs such as Beaumaris, East Bentleigh, Wheelers Hill or Templestowe.

The other benefit of the cheaper suburbs is that they have less natural or manmade heritage; NIMBYs shed tears over Camberwell but the locals of St Albans, Laverton or Keilor Plains are likely to accept and even embrace development especially if there are other benefits such as improved retail, transport and recreation facilities.