Our Top 50 Bus Routes - Part 2

Part 1 compared bus service levels by number of services per day. Such statistics are essential for scheduling and rostering purposes but have little meaning for the passenger. They are, in bureaucratic jargon, an 'output'.

Maximum waiting times, or service frequency, is far more useful for passengers. These are shown for each day of the week underneath the graphs in tables below:

Both the graphs and frequency tables allow some interesting comparisons.

The university-oriented feeder routes (401, 630 and 703) tended to have low ratios of weekend running. This is caused by the frequent weekday service and no or local-style weekend schedules. Fairly low weekend ratios are also found on outer eastern routes like 789/790/791 and the 888/889 SmartBus. Weekend services are particularly limited in the Manningham area, where service is concentrated on weekday commuters.

Trams are generally more frequent than trains, with the greatest differences during off-peak, evenings and weekends. This explains why the Knox Transit Link 732 (which connects to Tram 75) has a higher weekend service ratio than 571 and 896, which both feed trains.

Routes where weekend service levels are most similar to weekdays include 781/4/5 (where the basic combined frequency is 20 minutes 7 days a week) and 223, where only the Sunday timetable is significantly less.

Weekend services are important for some people but not for others. Good 7-day service makes the choice to live without a car more practical for more people, increasing loyalty and patronage. As an example, a change from 12 to 9 services per hour during a peak period would probably not dramatically change patronage (unless capacity limited) but the same change (from 4 to 1 trips per hour) on a weekend will likely influence a decision to drive. In other words, if you did not wish to buy a car, one would prefer a continually frequent route like Route 223 to one that offered a more 'peaky' service (even if more buses ran per week).

Another interesting pattern is the difference between Saturday and Sunday service frequency. Again this varies from small to great. Small differences are common with previously low-service routes where timetables were upgraded to a flat hourly frequency for each route for both days. Examples include the constituent routes of 781/784/785 and 832/833.

Large Sunday to Saturday ratios have two causes – either very low Sunday service levels or unusually high Saturday service levels that depress the Sunday ratio. An example of the former is 201/202/302, which has a sparse Sunday service compared to similarly ranked routes (refer 11-20 graph).

The author has observed that Sunday trading in the older northern and western suburban shopping strips is less prevalent than in the inner southern, eastern and bayside suburbs. Coversely the 'Saturday morning rush' tradition has probably also prevailed longer in the non-Sunday areas (in contrast Saturday afternoons are busier at 'destination' shopping malls).

Either it is demand driven by such patronage or a historical accident, but it remains true that Saturday daytime bus frequencies in the type of northern and western suburbs mentioned above approach weekday services and even trump SmartBus service levels. In all these areas Sunday buses were almost unknown until 2 or 3 years ago, when most routes got an approximately hourly service.

Examples of local routes that have a comparatively intensive Saturday service include 216/219, 220, 223, 406, 465, 467, 552 and 781/784/785. And to possibly confirm the above, all but one are in the northern or western suburbs.

Part 3 will present the frequency tables above in a different order and conclude with the benefits that combined frequency information can provide to both passengers and planners.

Our Top 50 Bus Routes - Part 1

Which are Melbourne's very highest service bus routes or route 'families'? Many would rattle off some SmartBus routes, but with one exception this isn't quite right. For there is a cluster of local bus routes that, through history and continued patronage, offer almost tram-like service levels seven days per week.

Service levels can be measured in several ways, including counting the number of services run per day or week or examining timetables for service frequencies. In this part I did the former, and the result appears below.

The first graph shows the number of services run per week for the 50 most frequent bus routes in Melbourne. In this and other graphs the total number of services is documented, even though some might be short workings and not serve all stops. I have also grouped routes where they form a 'frequent service corridor'. Others may use different groupings of related routes, so rankings of service levels may vary slightly. Click the graph to enlarge if required.

The SmartBuses are circled in orange. Blue and green circled routes are special services that connect with trains and trams respectively.

Route 732, the Knox Transit Link, has more services than any other bus route. An intensive service is provided on the portion between Vermont South and Knox City; the rest of the route offers a standard local-type service. Its high frequency is because it connects with every tram and trams run more often than either trains or buses.

Second is Route 401. This runs five days a week over a limited (daytime) span. However its 3-6 minute frequency is enough to give it its large number of services per week.

Third is a cluster of related routes towards Manningham. Fourth is the first SmartBus - the new 903. 903 (and other routes included like 703 and 732) have many short workings. All have been counted, so actual service levels along a particular section of route may be less than stated here.

The number of services running each day of the week is shown below. Large variations, both between weekdays and weekends and between Saturdays and Sundays are visible.

Part 2 will include further graphs and compare service frequencies, which is more useful for both passengers and transit marketers.

Contracts Contracts Contracts

The last couple of days have been big ones for train tram and bus operator contracts. There's enough about trains and trams in the papers and on the DoT's website. But in case you haven't heard, both incumbent operators have lost to MTM (trains) and KDR (trams), who now have 'preferred bidder' status.

What's this about buses, you may ask? Buses have a lower profile than trams and a much lower profile than trains. Similarly, their operating contracts (negotiated a year or so back) are not widely known and have attracted little attention. Bus contracts have always been an arcane topic; one would have to go back to the acrimony of the late 1980s (bus contracts dispute) or the early 1990s (franchising of Met Bus routes to National Bus and Melbourne Bus Link) to find the last time they were prominent.

A report by the Victorian Auditor General's Office sheds light on Melbourne's new bus contracts, their drafting and the role of the Department of Transport. It is recommended reading, and can be downloaded here.

What happens if contracts don't work out? It varies. When National Express walked in 2003 their operations automatically reverted to government operation until the metropolitan portion was relet to Connex the following year. It was a bit different with the regional intrastate network. where the government 'bought back' the rights it had sold to a lessee who couldn't make it pay. Read all about it in another recent report from VAGO.

What should we do about the 703?

Human Transit has done it again. Written a post that sparks one here. Jarrett Walker's post discusses transport service projects that expand yet dilute, potentially undermining their rationale and key selling point.

Here in Melbourne the opposite has happened with SmartBus. As its network spread the basic service level offered improved. All of the recently introduced SmartBus routes have better services than the initial pilot routes (703 and 888/889).

Route 902, the Green Orbital, will take care of the 888/889 when it starts next year. However nothing has been said about the future of Route 703, the other SmartBus pilot.

Route 703's key trip generator is Monash University Clayton, which it links to numerous Zone 2 trains and buses. It also has other transport interchanges and trip generators at Brighton, Bentleigh, Clayton, Sydndal, Forest Hill Chase and Blackburn.

A 15-minute service provides a SmartBus frequency on weekdays. Saturday buses run a 30 minute headway. The Sunday frequency is 45 minutes, with the Brighton end of the route receiving no service. Evening service mostly ends before the 9pm 'minimum standards' finish for local routes. The main exception is the eastern portion of the route, which sees hourly buses until about 11pm on weeknights.

A Sunday schedule runs on all public holidays. This is different to the new local route and SmartBus standard, which prescribes a Saturday holiday timetable except for Good Friday and Christmas Day. Partially balancing this are some extra services run on those public holidays that are university teaching days.

Hence Route 703's service level is best described as a hybrid between a regular bus and a full SmartBus route. Its weekday frequency equals or beats all metropolitan local routes bar two. The 11pm service on part of the route is almost as rare. On the debit side its weekend span and holiday arrangements do not meet local bus service standards, let alone SmartBus standards. Having said that, its 30 or 45 minute weekend frequency is roughly similar to the better local routes and the other pilot SmartBus (888/889).

How does one explain 703's 'limbo' status to the general public?

The answer is 'with great difficulty', especially when trying to market other SmartBus routes.

I can see three options for Route 703, as follows:

(a) Upgrade service to full SmartBus standards, though it doesn't fit in with any proposed orbital
(b) Keep its current timetable (or a slightly improved one meeting 'minimum standards'). Remove reference to SmartBus, instead treating it as a strong local route with a good weekday service
(c) Do nothing

(a) is obviously ideal. It would require additional funding, but is probably justified by its patronage. Maybe the university could chip in, given its revenue from international students (a key passenger group) and the opportunity cost of providing parking?

(b) is cheaper, incurring modest additional running costs for slightly improved spans (2 hours extra per week, plus public holidays and full route running on Sundays). The SmartBus totems could remain, but the service might no longer be SmartBus branded.

(c) is perhaps the worst option, unless you think that keeping SmartBus branding would encourage (a) above to be implemented. But if this doesn't happen, letting 703 limp along as a 'quasi-SmartBus' with its sub local route weekend span devalues the standing, meaning and value of all other SmartBus routes, due to 703's lower service levels and the need to keep saying 'except 703' when selling SmartBus benefits.

If nothing is done 703 risks being left behind as SmartBuses' unloved parent; one of the routes that started it all and proved it worked, but upstaged by its 900, 901 and 903 offspring.

For the good of the SmartBus concept specifically, and the bus network generally, a decision of whether to nurture or end Route 703's SmartBus status would seem desirable, with the urgency increasing as the Green Orbital start approaches.

What is a 'connecting service'?

The back of Melbourne train timetable booklets have a handy section on 'connecting' buses. This lists bus routes departing from each station and their destinations. This and Human Transit's post on transfers led me to ponder what qualifies as a connecting service.

At least in Melbourne, the term 'connecting service' is used loosely and devalues the term's potential meaning to passengers. It could be anything from an infrequent bus that departs from a busy road across from the station to a cross-platform transfer to a waiting train. In my view a genuine 'connecting service' imposes two main requirements; one physical access and the other temporal.

The access requirement means being able to get to the stop of the service that you are transferring to. If you can then the services can be described as 'intersecting'. Melbourne does this quite well; apart from a couple of cases (eg Southland and Campbellfield) if your bus goes near a railway you can be fairly sure there will be a station to change to not far away.

The time requirement is less definite and a judgement may have to be made, depending on mode and context. A 28 minute 'connection' is too short to be reliable at an international airport but is excessive for a suburban train-bus connection. However it might be just right for a connection between interurban trains at Southern Cross.

The best scheduled connections involve a bus for every train (eg 571 and 896 TrainLink services). This implies buses departing at the train service frequency ie harmonised headways. Acceptable compromises for local routes may include shorter operating hours for buses and their frequency being set to meet only every second train. However headways must first be harmonised (a bus every 30 minutes cannot reliably connect with a train every 20 minutes so does not qualify) with departures tweaked to minimise waiting. If it meets these requirements, I would describe the bus as 'connecting'.

Then there is what happens if the train is late. Are connecting services held for the train? And if so, is there a cut off time when the bus departs even if the train has not arrived, since buses have their own schedules and passengers to serve? If the answer to the first question is yes, and the policies on the second question are conveyed to passengers, then this could be defined as a 'guaranteed connection', although the guarantee could be described either as 'conditional' or 'absolute'. In Melbourne off-peak Alamein and Williamstown lines offer guaranteed connections as most of their passengers will be arriving on ex-city trains that continue to Ringwood or Werribee. A similar arrangement applies for tram routes 5 and 64 at some times of the day and probably country coaches (especially those that substitute for closed country branch lines).

To recap, there are three levels of service connectivity.

* The first is 'intersecting services', where passengers can transfer between services but no warranty can be given regarding reasonable waiting times or even if the service is still running. This applies to most buses listed as 'connecting' in the back of the train timetable.

* Secondly are the services that connect with short and predictable waiting times if both are to schedule. These require either frequent service or harmonised headways and can be regarded as genuine 'connecting services'.

* Thirdly is the highest and best level of connection, where services are held if there is late running. These are 'guaranteed connections', though the guarantee can either be 'conditional' or 'absolute'.

It would help if we used the correct term for what we mean to describe the relationship between a railway line and a passing bus route. Then terms like 'connecting service' get a firm meaning, and can become a selling point, much like Adelaide has achieved with 'Go Zone' and we have partially achieved with 'SmartBus'.