Tuesday, October 19, 2021

Timetable Tuesday #138: Melbourne's 'Rapid Running' bus trial - can it work?


One of the initiatives in Victoria's Bus Plan, released back in June (and reviewed here), was a 'rapid running' trial on certain bus routes. 

The Minister's foreword states:

Rapid Running is being successfully trialled on Route 246 and up to 10 more routes will be trialled before the end of the year.

So what is 'rapid running'? 

It's a jazzier term for what is more often called 'headway management' or 'frequency timetables'. 

Why would you do it? 

Consider a passenger making a short trip on a bus from A to B. They board the bus and, a short time later, the bus stops. Apparently it's running early and needs to wait for the timetable to catch up. If traffic conditions are light or scheduling is loose the bus could easily sit for five minutes while it waits for its scheduled departure time. 


That's not a good experience as the wait is a large proportion of the total travel time. And time not moving is perceived as longer than time in motion, even if slowly, so it seems worse. 

'Rapid running' aims to keep the bus moving for more of the time. It shifts the emphasis from exact times to maintaining a specified headway so buses no longer need to stop if ahead of time. Passengers making short trips like A to B might find them quicker as time point waits are removed. 


The three highs of 'rapid running'

Headway management is potentially useful on busy roads where varying traffic conditions greatly affect bus run times. Such variability makes it difficult to write a timetable that will be adhered to.  

However at least three conditions must be met for the trial to succeed. These include: 

1. High Frequency. You would never headway manage routes not frequent enough to be regarded by users as 'turn up and go'. Infrequent buses need fixed timetables, preferably timed with the trains they feed (especially if these are also infrequent). Waiting time is perceived as longer than time in the bus (even if stopped at a time point) so unless there's enough buses to run short headways, 'rapid running' is more trouble than it's worth.  

How frequent must a bus be before you'd consider it suitable for headway management? Expert opinions vary but there seems agreement that you not do it unless services were every 10 minutes or better.  


Based on a 10 minute cut-off, where are Melbourne's frequent bus routes? 

Unlike other cities of our size we have only a handful, even if you look at weekday timetables only. They're labelled below.  

Frequent corridors formed by multiple routes are harder to rapidly run (at least in both directions) so are shown but not labelled. 


Our lack of frequent routes makes finding even 'up to the ten more' suggested in the Minister's Bus Plan preface a tough ask. Now on to the remaining two conditions. 

2. High technical, operational and cultural readiness on the part of the bus operator and contracting agency. A quality 'rapid running' operation with even intervals between buses requires more close monitoring and responding via a control centre using traffic data from the road management agency. That can then be used to provide information feedback to drivers and an ability to vary the number of buses on a route to maintain headway under busy or quiet traffic conditions. Passenger counters on buses might enable better responsiveness to surges in patronage, but again there needs to be the ability and resources to respond quickly.

Ideally bus data would also flow back the other way, to the traffic authority, so that traffic signal phasing gives buses a good run. This data revolution has implications for bus management practices from depot locations to staff rostering, operator payments and performance monitoring, management and penalties.  

3. High passenger trust. Take away a timetable with specific times and passengers will scream. Especially if it's a low-trust system where transport departments are seen as distant from passenger concerns or not really in control. The following can lower trust in a bus network: 

a. Services are run by poor quality operators (like the soon to exit Transdev Melbourne about 4 or 5 years ago with its dirty, cancelled and unmaintained buses) without effective oversight from transport bureaucrats (who can't be too tough in case they damage their prospects for jobs elsewhere in the industry, including at those they previously regulated). 

b. 'Rapid running' is introduced simultaneously with unpopular changes like privatisation and radical network reform, especially if it removes coverage (like what was proposed then scrapped in Adelaide last year). It is better to introduce it separately and gradually, like we are doing. And explain it well.  

c. Real-time information is either missing or unreliable. Seeing a bus show up as coming, not arrive and then pass after you have given up waiting really shakes faith in the network. Real time information systems need to be at least an order of magnitude more reliable than the system they're monitoring to win user trust. For buses in Melbourne we're not there yet! 

d. There is no public operational performance reporting or accountability. For a long time we reported train, tram but not bus service delivery in Track Record. Buses with fixed timetables are now reported too. However there still needs to be a way to report performance of 'rapid running' bus routes. For a route with a nominal headway of 10 minutes, reporting could highlight instances where fewer than 6 trips per hour run and gaps between buses exceed (say) 12 minutes. 

e. There is a passive official acceptance that bus passengers come near last in the transport system, with few exclusive lanes and services cancelled or diverted when car traffic volumes prevent buses from moving. 


Another example of bus' low status is what happens if services are disrupted. If Metro train services are suspended replacement buses are brought in. Indeed contingency arrangements exist with bus companies to do just that. Whereas if a bus operator can't run its buses then there are rarely arrangements for others to take over, so trips just don't run with 30 to 60 minute gaps, even on major routes. The main recent exception was during Transdev's fleet maintenance crisis of 2017 where other operators' buses stepped in to run Transdev routes. 

'Rapid running' could be described as a 'social contract' between bus operator and passenger. We take away your timetables in exchange for running a frequent service and better responding to traffic conditions. Its acceptance depends on passenger trust that needs to be earned beforehand.  

I think it would be fair to say that generally Melbourne has a low to moderate trust bus system, although parts, like the most university shuttles, have much higher reputations. That's important because if trust is high it is likely easier to win (even if grudging) acceptance of bus reform like but not limited to 'rapid running'. 

Which routes are getting rapid running?

The first route we knew about to get rapid running was the 246 down Punt Rd/Hoddle St. This is a popular route along one of inner-Melbourne's busiest road. It is one of just two 7 day routes to feature a 10 minute interpeak service (the other is the 402 across from Footscray). 

'Rapid Running' for the 246 commenced in March 2021, before the Bus Plan came out. 

There were enough results by June for the minister to be able to say it had been a success. 

A little later the PTV website page advising the 246 trial was updated with other routes listed for 'testing frequency timetables'. These were described as 'selected high frequency bus routes'. 

Timetables would be removed from stops but the buses would continue to follow timetables (unlike the 246). Passengers would be surveyed on what they thought, with the possibility raised that the routes could change to a 'turn up and go' service like the 246. 


Let's go through these routes, starting with the most frequent first. 

* The 401 Melbourne and 601 Monash are both university shuttles that are already very frequent (eg up to every 3 or 4 min). All users would definitely be using them as a turn-up-and go service right now. These are exactly the sort of services that you'd have frequency timetables for with their short lengths likely making management easier. 

* 301 is another university shuttle (Reservoir - La Trobe). This runs every 10 minutes. However the train service it was designed to connect to at Reservoir is only every 20 minutes off-peak. That makes arrival time important. Even if the bus is frequent you still need to have scheduled times and fixed timetables to avoid the possibility of just missed trains. Hence I'm wary about 'rapid running' on this route despite its frequency.   

* 402 is the only other single seven day route in Melbourne that's every 10 minutes. Like the 246 it runs seven days but less frequently on weekends and with shorter operating hours. And it can similarly be described as being of moderate length, unlike the short university shuttles discussed thus far. Length is important because in a few lucky cases there may be times where you might be able to save a bus or boost frequency if run times are consistently shorter than the timetable allows. This makes it the next  most suitable route for a trial after the 246. 

* 250/251 are both every 20 minutes each. These provide a 10 minute combined service through Melbourne's inner north. This is why only southbound trips would have their timetables removed. This different presentation of information may confuse passengers since northbound trips will retain timetables at stops and the same passengers will likely be using both.

* 513 to Glenroy. This absolutely should not be a part of the trial even though it serves busy Bell Street. Not only because it's a long and confusing route with alternating paths in the east. The main reason to object is it's just not frequent enough. The PTV list above claims 15 minute frequencies, but this is wrong; the 513 operates only every 20 minutes, a frequency that definitely needs a timetable. The 513 should be removed from the trial until the route is reformed and services upgraded. 

What wasn't mentioned? Most conspicuous is the other frequent university shuttle, the 202 from Parkville to Victoria Park. Its 10 minute frequency is like LaTrobe's 301 and, beyond Clifton Hill, the trains it connects to are every 20 minutes interpeak. It only started on September 20, so maybe they wanted to try a regular timetable first. 

Something that would have disrupted 'rapid running' trials on routes run by Transdev (including the 246) is that reduced driver availability due to COVID has seen mass cancellations with some services slashed by 50% or more. With low frequencies, even on some of our main bus routes, you absolutely have to have fixed timetables and 'rapid running' is out of the question. 

Other possibilities

If things go well and the trials were a success, where else would you consider it? 

If the lessened legibility of doing it in one direction only on the inner portion of a corridor of routes is acceptable then two other potential corridors include the 905/906/907 from Collingwood into the city and the 200/207 from Kew into the city. These provide approximately a combined 5 and 10 minute service respectively. 

Another possibility is peak service on the routes with intensive peaks eg some of the 900-series SmartBuses (900, part 903, 905, 906, 907, 908) and possibly others like the 220, 234 and 465. 900, with its two companies running the route, would be operationally difficult if not impossible. Peak buses serving Fishermans Bend might also be close to the 10 minute frequency required. 

Key to further expansion is Sydney-style bus network reform to simplify services and increase the number of routes operating every 10 minutes or better. This is especially if 'rapid running' can reduce dwell time so that headways can be got down to 8 or 9 minutes. It should be emphasised though that rapid running should never be regarded as a substitute for useful but politically difficult initiatives such as bus priority and segregation on key corridors. 

The sort of short to medium distance and high patronage potential routes that could justify, high priority, 5 - 10 minute frequencies and rapid running might include the likes of:  

* 201 Upgraded Box Hill - Deakin Uni shuttle
* 220 Megabus along Ballarat Rd between Sunshine, Footscray and CBD
* 406 Megabus between Footscray, VU and Highpoint (with extension & other reform)
* 733 SRL SmartBus between Box Hill, Mt Waverley, Monash, Clayton, Southland, Sandringham
* 904 Megabus between Coburg, Preston, Northland and Heidelberg
* 900 Megabus between Caulfield, Chadstone, Monash and Stud Park
* 907 Megabus between City, Doncaster and Mitcham

These included upgraded regular routes and or significant reform to other routes. This lowers the cost of providing the desired 10 minute or better service needed for 'rapid running'. This needs to be done before 'rapid running' is attempted as all are currently only every 15 - 30 minutes interpeak. 

More ambitious upgrades, again with new 10 min or better off-peak service, might include:

* New 620 SmartBus Caulfield - La Trobe University
* 901 SmartBus between Ringwood, Dandenong and Frankston
* 902 SmartBus between Melbourne Airport, Greensborough, Doncaster, Springvale and Chelsea
* 903 SmartBus between La Trobe Uni, Heidelberg, Doncaster, Box Hill, Chadstone and Mentone
* New Wollert - Northern Hospital - Epping - Plenty Rd bus wormhole

These are mostly new (620, Wollert) or realigned (902, 903) high patronage potential bus routes. Current frequencies on these corridors are from 15 to 40 minutes so upgrades would involve significant extra resources. However some would be clawed back by reducing route overlaps, particularly in north-eastern Melbourne, where frequencies are out of kilter with usage. Also reform would provide a substantial BRT style network for much of middle and outer Melbourne. The success of these longer routes (especially the split orbitals) likely depends more on the priority that can be accorded (to enable faster and more reliable timetables) than band-aids like 'rapid running' alone. 

Conclusion 

Rapid running is a 'big city' bus idea. There may be a case for it on some routes in Melbourne. However we don't have enough of a frequent 'big city' bus network to make it widely applicable.  

A much larger scale rapid-running roll-out requires us to develop a frequent 'top tier' bus network first. Several high patronage potential corridors have been suggested for this. 

'Rapid running' is not a cure-all. The real game-changers for buses are simplified, more frequent networks and 7-day bus priority to enable shorter and more reliable journey times. This would increase passenger boardings per kilometre and free up resources for further service upgrades and frequency improvements.

Your thoughts on all this are appreciated. Have you tried the 246 or other routes lately? What have been your experiences? Would you rather faster time in the bus or set timetables at stops? Or maybe you think some routes should or should not adopt rapid running in preference to those discussed. Please leave any comments below. 



5 comments:

Anonymous said...

I've seen some bunching of 246 buses through Elwood. I have in the past before the changes I have spent five minutes in a stationary 246 at Victoria Parade waiting for time. There will be always be one driver who is faster than another and so the service does need very close monitoring to ensure headways are even. I think that is where the system will fall down.

Daniel said...

The word "trial" is very important here. I've heard repeatedly that the trial is genuine - it's not going to become permanent by default.

The reason routes with lower frequencies have been included in the trial is that they want to test if they work with rapid running. My bet is they won't; 10 min frequencies will be okay (the feedback for 246 has been good). 12-15 mins maybe (especially on routes which are 10 for much of the time).

20? I really doubt it. But remember that 513 is only every 20 off-peak, but does run every 15 minutes in peak, eg first service until about 9am, then again from about 2:30pm to 7pm... so it's a fair bit of the day, and the PTV page indicates that's what they're targeting.

As long as they're true to their word and do a proper evaluation passenger reactions to the trial, then it's fair enough to try it, to establish the boundaries for a later broader rollout.

On routes 250/251 doing rapid running in one direction, the question is how citybound passengers use the 10-minute frequency now. Do they inspect the timetables? Do they just go to the stop and/or use the real-time information? Maybe it depends if they're making a trip where either route will do, or where they need a specific route. Again, a trial is how you figure out if it's a positive or negative for passengers.

Tramologist said...

Rapid running routes also have fixed departure times at origin. The only difference is the removal of intermediate timing points. The university express shuttles have no intermediate stops and therefore don't have intermediate timing points anyway, so it's pointless to include them as rapid running routes.

Marcus W said...

For route 402 a big gotcha with keeping even headways is that it passes through two level crossings - Craigieburn and Upfield lines, both located on Macaulay Road. In peak times, this could lead to large gaps between buses, but then again, it's a problem right now with timetabled serivces.

Anonymous said...

The rapid running 246 is great in terms of frequency but the fact that it loses the bus lane is a killer.