Building Melbourne's Useful Network Part 90: Getting the most from Keysborough South's second bus route

It was promised by local MP (and one-time transport minister) Martin Pakula before the 2018 election. It got funded in the May 2019 state budget. Even leaving aside the fact that it takes yonks to set one up in Melbourne, you'd have thought we'd have heard more by now. Even if if was just some public consultation or a start date announcement. But we haven't. 

What am I talking about? A bus route promised for Keysborough South. It's a growth area that's been crying out for buses. Its first bus started several years ago when the 709 was extended from its old Waterways terminus to Noble Park via Keysborough South. This crosses but doesn't go along Chapel Rd. Hence the call for a second route. 

Other new bus routes take just over two years from budget funding to implementation. How much is Keysborough South's lagging? Assuming there is no consultation it's slower by about a year. However consultation on its exact route was promised. Even discounting COVID that adds extra time with implementation about a year after consultation. And we are not yet at the consultation stage (if it's happening) for Keysborough South. 

Have any clues been given as to where the new route might go? Pakula's promise is that "this new service will travel along Chapel Road, connecting locals to the surrounding schools and shopping centres". Some of these are identified on the map below (click for better view). 

A base case (with standard 'grafted on planning')

We don't know what the Department of Transport has in mind for the new service. But I've plotted a base case option below. It serves Chapel Rd, shopping centres at Parkmore and Keysborough South plus some schools. It also connects the area to its nearest station at Noble Park. Hence it meets the promises made for it. 

Its length is also efficient. Melbourne buses operate at an average speed of about 22 km/h. Because it is just under 10 km long, one bus is sufficient to provide an hourly service with recovery time at the end. And two buses will comfortably provide a half-hourly service, like that which runs on the relatively new 627 and 709 routes. I will assume that the new route will also run at this frequency.  

While this fix meets a political promise and need not change things for users of other routes, it has shortcomings. For example trips to Parkmore from Chapel Rd are indirect as you are going south before going north.  The same applies for Perry Rd if going to Noble Park station, with the complication that  you would be waiting on the opposite side of the road than for the existing 709, which also goes there. As a distributor road Perry Rd is likely to carry significant traffic. It may present hazards if passengers see a bus they want on the other side and cross hastily without looking. 

As is common Department of Transport planning practice (including with the 627 and 709), I've assumed no changes to the existing bus routes that the new one overlaps. This is in keeping with  their weakness in challenging the opportunity costs of lazy 'grafted on' approaches to bus network planning. 

I don't think this is necessarily all due to a lack of planning expertise in the DoT. Instead there appear to be internal blockages somewhere high in the chain of hierarchy that stops or slows good things being done. Even where the government and/or minister may be risk averse with regard to bus network reform, the case should at least be made (though the present minister has indicated support for a bolder approach to bus reform). 

In another context (a major IT project failure) the state auditor-general recently said that "(the Victorian public service) must also re-examine its approach to providing robust advice to the government, especially when the government may not be receptive to the advice". It is possible that such a recommendation may also have benefited past and present DoT executives who have been unable to win ministerial support for bus network reform.

  

Network options (with a wider network overhaul)

I discussed a simplified Greater Dandenong bus network back in Useful Network 4 . Something mentioned then was the large number of infrequent routes (typically hourly) with overlapping coverage. Nothing has changed since. 

However the additional resources (likely two buses) for the Chapel Rd route could go a fair way towards funding a 2014 Brimbank style network revamp for Springvale, Noble Park and Keysborough with wider benefits like simpler and more frequent services. Had such planning begun as soon as the 2019 budget come out we would not be far off seeing that network being implemented now. 

Areas where you could simplify and improve services are below. Because many routes eat into each other's catchments, it should be possible to get some frequency upgrades with similar or improved coverage. 

If an hourly 10 km long route requires one bus to run then saving 1 km from an hourly route should require 0.1 bus less. Route level calculations are coarse so trimming fractions of buses does not necessarily save much. But with interlined schedules (where one bus might run several routes in a shift with layovers minimised) it becomes finer, tending towards the theoretical figure on a network level. This is particularly for larger operators such as Ventura (who runs all non-SmartBus Keysborough area routes). 

Potential km savings by section of route removed are mapped below. 

812 Kingsclere: 1.5km* (would only do if a route added to Corrigan Rd to serve stops)

812 Dandenong - Keysborough 6km (would only do if nearby 813 upgraded)

815 Dandenong - Keysborough 6km (would only do if nearby 813 upgraded)

815 Keysborough - Noble Park 3.5 km (portion would gain trips with revised 815 - see later)

(*) The distance cut is greater as it includes Corrigan Rd as well as Kinsclere Av. However this is partly offset by rerouting via Cheltenham Rd instead. The 1.5 km figure reflects an overall saving. 

The total is approximately 17 km. Given that these are all served by hourly services the resources of approximately 1.7 buses is in play. There is no significant loss of coverage and very few stops would be left without service. Thought may however need to be given to better serve the area around Railway Parade Dandenong depending on usage of existing stops. A diversion of one of many routes in the area would achieve this. 

Option 2

This network overhaul not only delivers a Chapel Rd service but also provides a third Keysborough South route, simplifies other routes and increases frequencies across a wide area from Springvale to Dandenong south of the railway. The savings above make it economical to deliver. Yet coverage is expanded and very few stops would have service removed. 

Key components include: 

1. A Route 813 that has its frequency doubled for the 13.3 km between Dandenong and at least Springvale. This is funded by consolidating the Dandenong - Keysborough portions of the 812 and 815 into an upgraded 813 service. While this reduces service on some portions from three to two buses per  hour, the revised timetable features simpler and evenly spaced trips 7 days per week over longer hours (eg the current 815 only operates 5 1/2 days per week). The 813 upgrade would mean a high patronage portion of Springvale gains doubled service  and a busy portion of the 902 SmartBus would get some relief. 

The main complication is that 813's service north of Springvale station would likely remain at hourly unless further network reform takes place. A potential option could involve terminating 813 at Springvale Station with the 813 portion north of Springvale Station to be run as a shortened 814 improved to every 30 to 40 minutes (instead of the current 60  minutes). Springvale South could have a rerouted 811 instead of the 814 to gain 7 day service. Any resource shortfalls could be met by pruning frequency or hours on the now overserviced 704 which is much less used.  

2. A new coscheduled pair of routes (815/816) serving Keysborough South including Chapel Rd. These are about as direct as can be, starting at Noble Park Station. It would give Corrigan Rd a direct bus to its nearest station it doesn't currently have. 816's southern terminus could be the new Keysborough South shopping centre. However services would be planned (as with Point Cook's very popular 494/495) so that buses there continue via a revised Route 815 to Noble Park via Parkmore Shopping Centre. This allows people in a wide area to have a single bus trip to and from Parkmore. Also if departures are offset those who just miss the 815 at Noble Park Station have the option of catching the 816 and going the other way around (and vice versa). As part of this change to improve directness  and save kilometres Route 812 would be rerouted to go straight to Parkmore instead of indirectly via Corrigan Rd (that would gain the more frequent 816). 

The round distance for this pair of routes is about 13.3 km. That should allow a 40 minute frequency with the two buses assumed. While this is less than the 30 minute frequency for the single Keysborough South route in Option 1, scope exists to offset departure times at Noble Park to assist those near both 815 and 816. Also the frequency and operating hours is superior to the current 815 in the Noble Park area, delivering a substantial improvement across a wider area than just Keysborough South. Legibility is also improved, with passengers going in the expected direction to Noble Park Station and some being able to take advantage of same direction buses on the 709. 

3. Optional extension of 812 to Dandenong via Dandenong South. This could improve connectivity to jobs from the Keysborough area.  It requires additional route kilometres so involves additional expense. Although it is an hourly service it runs 7 days per week which is more than can be said for the 857 in the area. Hence it may be possible route route the 812 via the residential area instead of the 857 to provide homes this expanded service. 

Routes 709 and 824 in the area could stay the same though it is desirable that 824's frequency and operating hours are improved. Having said that, many near the 824's catchment will benefit from the new 816 and the upgraded 813. 

While a large improvement over what's there now, this network is not perfect. The 813 should ideally be every 20 minutes given the favourable demograpics for buses in the area it serves. And the 815/816 pair should have a higher frequency than the resourcing allows.

Another deficiency is that routes are still too close together in the Noble Park area. However economising by terminating 709 at Keysborough will remove its train feeder function. I would be very wary about doing this as the 709 is quite a good performing route. Another possibility could be to run the 815 and the 709 on Bloomfield Rd. If the 815/816 pair were boosted to every 30 minutes and the 709 and 815 staggered at Noble Park then the result would be a 15 minute combined service as far as Westwood Bvd (Keysborough South) - ie superior to anything now in the area. People near Buckley St would need to walk about 300 metres east to Bloomfield though some would benefit from the new 816 on Corrigan Rd (300 metres west). Overall this could be a 'stage two' network reform to be considered after the first changes have had a chance to be tested. 

Option 3

This is a more austere network that delivers a new route down Chapel Rd but not via Corrigan Rd. It enables higher frequencies but at the price of somewhat less directness.  Key components include: 

1. A doubled frequency Route 813 as per Option 2. This would again be paid for by merging the Dandenong ends of 812 and 815 into it. 

2. A Keysborough South route comprising a single linear route from Noble Park. This would operate as a modified Route 815 extending there instead of Dandenong. This route is 8km long and could comfortably operate every 30 minutes with two buses. Such a 30 minute frequency could provide more even connections with intersecting routes like the 709 or the upgraded 813 (both of which would run every 30 minutes). 

Route 709 and 824 would stay the same. Because there is no dedicated Corrigan Rd route the 812 would also remain indirect, going along parts of it. However it would still finish at Parkmore. 

The extended 815 has a weak terminus near Mt Hira College. Also its 8km length is not very efficient for a 30 minute frequency (whereas a length near 10km would be). Scope exists to reduce its length to under 7 km by terminating it at the Keysborough South Shopping Centre. This might (just) allow a 20 minute peak period service to be run with two buses. Meanwhile one bus could run a 40 minute off-peak and weekend service if economies were needed. 

Conversely, if connections to the Dandenong South industrial area were desired the route could be extended east along Greens Rd to Dandenong. Extra buses would however be required unless peak frequency was greatly reduced. Better Dandenong South public transport connectivity has long been advocated for by the City of Greater Dandenong (such as in this 2003 era strategy).  

Conclusion

Keysborough south is due for a new bus route soon. If it was anything like the last additional route (the 709) and others in the south-east (like the 627) it will be plonked on to an existing unchanged network. 

While expedient this carries significant opportunity costs. 

Described are two revised networks that deliver benefits wider than just Chapel Rd across several suburbs. And they should be fairly cost-effective too. 

I'd like your thoughts on which of the three networks you prefer. Or maybe there's another variant I haven't mentioned. Please leave your ideas in the comments below. 

See other Building Melbourne's Useful Network items here