Monday, April 28, 2008

Why fast interchange matters

a response to a Railpage discussion about Box Hill and the importance of minimising bus-train interchange time

There's worse interchanges than Box Hill, but best practice is that train to bus interchange should be maybe 30 - 60 seconds via just one ramp, steps or escalators (eg Werribee, Boronia or newer Perth suburban stations).

And it should not be possible to exit any suburban railway station without almost bumping into a bus timetable or information - something that is denied to Box Hill passengers.

Every minute longer in bus-train access times lengthens bus standing times by 2 minutes, assuming a co-ordinated pulse timetable system. This reduces the capacity of bus interchange and is poor efficiency. Where stations are midway along a bus route (and not at the end like Box Hill) long bus dwell times can also increase journey time for through passengers.

Two 'perfect connection' examples, both assuming a 4 minute buffer to allow for late trains/buses.

* 1 minute station - bus access time (ie good design)

Bus arrives: 9:55am
Access time: 1 min
Wait time: 4 min
Train Arrives/Departs 10:00am
Access time: 1 min
Wait time: 4 min
Bus departs 10:05am

(min bus dwell time 10 min)

* 3 minute station - bus access time (ie poor design)

Bus arrives: 9:53am
Access time: 3 min
Wait time: 4 min
Train Arrives/Departs 10:00am
Access time: 3 min
Wait time: 4 min
Bus departs 10:07am

(min bus dwell time 14 min)

The general concept is better explained in the diagram below:

Time is money. Multiply that extra 4 minutes by the number of bus movements per day, and then per year. It's big bikkies! Or try to economise by skimping on connections, but since that may add 30 minutes to passenger travel times it's a false economy as they'll drive instead.

In short, each second counts, and no effort should be spared in improving interchange, whether the delays be caused by poor initial design (Box Hill), a botched redevelopment (Melbourne Central) or placing car ahead of pedestrian access (most places - Caulfield example below).

Sunday, April 27, 2008

New suburban train timetable adds 105 weekly services and evening capacity

A new Connex suburban train timetable started today. It added 105 weekly services and ran 6 carriage trains up to at least 7pm on all lines. Stony Point also got an upgraded train and timetable.

Below is a detailed item on the changes. It foreshadows further changes in November 2008 to 'untangle the network' by running peak Werribee trains direct to Flinders Street and operating Epping and Hurstbridge trains clockwise through the City Loop all day (instead of reversing at midday and going counter-clockwise in the afternoon). 

https://web.archive.org/web/20080416015740/http://www.connexmelbourne.com.au/news.php?newsid=243&g=Array

Timetable aims also summarised in this Eddington report submission:

https://web.archive.org/web/20080722032100/http://www.connexmelbourne.com.au/media/docs/Connex_VTA_Eddington_Study_Response_July_2008.pdf

Note: This item written retrospectively for posterity.

Friday, April 25, 2008

Service standards for buses

When people front up to a bus stop, they don't know what to expect.

For a start (unlike a tram stop or railway station) it might not have a timetable.

But assuming it does, the timetable can show a big variations in services. For instance, it may run frequently on weekdays but not at all on Sundays. Some services operate weeknights but not on Sundays, while other go Sundays but not weeknights. Regular routes may provide a superior service to SmartBuses on some days but not others. Public holidays and reduced December/January timetables are other causes of confusion as there are no network-wide standards.

Since the service levels of some bus routes can be low, it is easy to tar all with the same poor brush. And this is understandable given that even a single route can be simultaneously good or poor depending on the day of the week and time of the year.

The SmartBus program provides a premium and visible level of service on five routes. Currently they are all in Melbourne's east but eventually they will extend around the northern and western suburbs via proposed orbital routes. Except possibly for peak times, SmartBus provides a consistent level of service on weekdays. However weekends and evening SmartBus services still have a way to go, with some regular non-SmartBus routes offering superior service.

At the local level there are the minimum standards improvements which will give local routes at least a basic hourly service until 9pm and standard public holiday service arrangements. Over 200 of Melbourne's 300 bus routes are to be upgraded, and we are currently just under half way through.

When minimum standard upgrades are finished, it will be possible to talk about a minimum service standard of a bus at least every hour until 9pm. And given the 90%/400 metre rule, we could add it would be within a 5 minute walk of most homes as well.

Left out of this is a big body of busyish routes that are neither SmartBuses nor quiet local routes. They may serve shopping centres, universities or a major corridor. They may be highly patronised routes, such as the 200-series out to Sunshine or 250/251 to Northland/Latrobe University. Or they might be the only direct link for some suburbs missed by trains or trams (eg various City-Doncaster routes). Many of these already operate to quite high service standards, and even, as noted before, run more frequently and later than SmartBuses.

It is these intermediate/'nonSmart' routes that are currently getting a bad deal. They don't get the promotion or infrastructure of SmartBus services. Neither have they had the public holiday standardisation of local routes. Where these services comprise two or three similar routes, timetables may not always be in composite form, so Joe Public can't readily see the higher service level offered. Unlike SmartBuses or local routes they don't have documented service standards. To the passenger it's just another route number, and study of a timetable is needed before knowing whether the route is hourly until 7pm or every 20 minutes until midnight. These good but 'non-smart' routes will never reach their potential until this is addressed and their superior service is highlighted.

The solution to all this is contained in a media release quoting Bob Cameron, the Member for Bendigo West, when announcing upgraded local bus services in Bendigo.

The salient sentence is this:

"There will be services at least every 30 minutes on prime routes and every hour on secondary routes,” Mr Cameron said.

Since the hourly 9pm 'minimum standards' are just that - minimums - we can label all those routes that conform as 'secondary', just as Mr Cameron did for Bendigo's quieter routes.

SmartBuses are usually better than that, so it's logical that they are 'primary'. Ditto for the 'nonSmart' routes if their service level is high enough. Document these primary standards, make some minor improvements to ensure adherence and standardise public holiday arrangements and we've effectively doubled the SmartBus network for very little cost. After all, who really cares if a bus is 'smart' or otherwise if it's like Route 220 and runs every 15 minutes until after 11pm?

The few routes that conform to neither primary or secondary status (perhaps because they are peak-hour only or serve semi-rural areas), can be labelled 'tertiary'.

Once we have a proper primary/secondary/tertiary standard, it can be explained easily to passengers; just as the MP did in his one-liner. We can say things like: Primary routes have a bus at least every 15 minutes during the day and every 30 minutes until midnight and secondary routes are at least hourly until 9pm. This exercise now is impossible since current timetables have dozens of variations and requires hundreds of words to explain.

Primary service can be promoted at stops, in timetables and on websites similar to what Adelaide does with its 'Go Zones'. Local public transport maps at all railway stations and bus interchanges would show the primary routes as thicker lines. The remaining secondary and tertiary routes would be thinner lines.

By differentiating primary and secondary routes, and heavily promoting the former, we've created a quite different, more versatile and more usable bus network. And this should translate into higher bus patronage overall, with the potential of upgrading popular secondary routes to primary status.

And we've met the challenge set at the beginning, which was to ensure that passengers know what service to expect, just as they do now with trains and trams.