Sign at Frankston Flinders Rd

Those who have not recently visited my blog may not have seen the photo of the junction points at Baxter, that appear to have been replaced sometime after 1975 (and before 1981, when the line closed).

Of course, by 1975 it would have been clear to the powers that be that the line would close. It therefore seems remarkable that the points were swapped. I guess the heavy freighters running up from Long Island needed to charge Langwarrin Bank on heavier rail. However, straight-railing the section, and closing the branch, may not have been politically feasible at that point.

The Mornington Rail Preservation Society have an excellent tourist railway arrangement running between Moorooduc Hwy, Moorooduc, and Yuilles Rd, Mornington, one that I have often stated captures the feeling of the 1970s on the VR branchline network: The K Class with wooden cars (substituted by a T class at the last minute). The railmotors. The clickety clack of the short rails. The piling out from carriages onto unmade platforms.

The society have as their stated goal to restore back to Baxter when the time is right. When I visited this morning, the Frankston Flinders Rd level crossing stretch was in poor condition, with maturing pine trees growing between the rails. This is probably no worse than the Musk-Bullarto section of the Daylesford line, which also had mature trees removed from track before it was restored. In addition, while we know the junction consists of a solid turnout, the Baxter Station area is now very basic and would require a massive amount of work to reinstate a yard for a tourist railway. The obvious place to build the yard is on the opposite side of the Stony Point line from the branch but I doubt the freight operators would want a second (facing) turnout, and the worry of tourist trains crossing the main.

Contributors in other places have often suggested, I would say naively, that the Mornington Line be reopened for general passenger traffic.

The main problems with this have been that Mornington to Frankston (the first logical destination of a Mornington commuter) route by rail is very very indirect compared with road.

I will explain the topography briefly.

Because the twin peaks, so to speak, of Mount Eliza and Mount Martha effectively block a coast railway from south of Frankston to around Dromana, no railway was built south along the shoreline of Port Phillip Bay. While a railway did eventually reach Mornington, the approach to the Mornington Peninsula was to take the line south from Frankston but swing sharply to the East, to gain height more slowly to reach the crest at Langwarrin.

From there, the height was immediately lost as the line descends to Baxter. From Baxter, the topography is either flat or gently undulating until almost Mornington, where the current train terminus at Yuilles Rd is sited on a sizeable grade down into Mornington township on the coast.

This diversion via Baxter was less troubling for the ultimate destinations on Westernport Bay such as Hastings and Crib Point (and previously Balnarring and Red Hill South) the diversion was probably fatal to the Mornington service. In fact I would suggest it was probably the single factor in the line’s failure, as the area had already become urbanised when the line closed, was no less an origin of commuters than say Melton, and was ceteris parebus (all other things being equal) an excellent candidate destination for electrification.

However, as I have argued on my own blog in the discussion of loops (for example, my rant on the failure of the City Loop) diversions are usually most wasteful for the customer of the place at the end of the diversion, but not necessarily for those in the middle. Using the City Loop as an example, a customer going from Richmond to Flinders St will find the diversion via Melbourne Central completely wasteful (and the leg from Melbourne Central to Flagstaff is actually going at a 90 degree angle to where the customer actually wants to go).

And in this case, while the Mornington commuter will find the journey via Baxter completely wasteful, this waste diminishes as you head along the line through Moorooduc to Baxter itself. From Moorooduc for example, while Baxter is definitely the ‘wrong way’ to Frankston, the amount of diversion is much lower (I will look at this below).

What are the transport needs of the Mornington Peninsula?

My gut feel is that if there was no rail service down there already, you wouldn’t build one. But working on the premise that equity demands if one area has a good rail service, another area with the same attributes (population, distance, density and overall demand) should also have one, you would keep the Stony Point line. Freight justifies most the line, as far as nearly Bittern, and the balance is short enough to wear the expense.

The Eastern Peninsula (the Westernport Bay side) is therefore served, with two major towns of Hastings and Crib Point served, and only two towns (Flinders and Balnarring) missing out (Balnarring was once served by an infrequent branch line from Bittern).

The Western Peninsula (the Port Phillip Bay side) completely misses out, now that Mornington has no service.

Road transport provides for all the transport needs of the western side.

Compared with a hypothetical rail service, the bus service to the end of the Peninsula at Portsea is poor. Assuming a train followed the times of an average, pre-RFR train service on this hypothetical line, over a distance of 55km the train should be able to cover the distance in 60 minutes, including 8 stops (using Donnybrook to Seymour as the comparator).

The bus service, with typical transit times of 100 minutes from Young St to Hotham St net of many, many possible stops, is very poor in comparison. Adding a typical 60 minute train transit from Frankston to Melbourne results in a total transit time of at least 160 minutes, or 2 hours 40 minutes, which is very poor for a 100 km journey.

Adding the poor frequency, the need to change at Frankston, and the use of an urban rather than intercity bus, makes travelling the Peninsula to Melbourne by public transport is a poor prospect compared to say Ballarat or even Korumburra.

The Eastern side is somewhat better. Crib Point to Melbourne is 70 km, and about 90-100 minutes by rail, including a change at Frankston. The use of rail, rather than a road alternative, clearly improves the usefulness of public transport to the interurban customer. Even, as in this case, where the rail transport offered is somewhat unattractive (and could be significantly improved).

But rail will always be an over-the-top option for many areas, including the Mornington Peninsula. The fact that a legacy rail service exists to serve the eastern side will benefit that side and not require a major positive evaluation to keep this service going. The lack of a legacy rail service to the western side, even with the Mornington line down, will probably prevent rail ever being considered for this side of the Peninsula.

The recent opening of the Eastlink freeway to Frankston has reduced road travel times to the entrance to the Mornington Peninsula, but (as freeways often do) increased political and local pressure to build the Frankston Bypass to reach the existing Mornington Peninsula Freeway which finished between Safety Beach and Moorooduc. Completion of this section would provide continuous freeway from Melbourne to Rosebud.

The excellent service provided by these roads to motorists is better on the Western side than the Eastern side. The Western Port Highway, which serves the Eastern side, leaves the freeway at Lyndhurst, and steadily deteriorates as it heads south. First the proliferation of round abouts which slow the motorist down and lead to jams; then the road becomes single carriageway around Langwarrin; finally the curvature increases as you reach Hastings and have to weave in and around the townships down there.

One interesting consideration is that the Mornington Peninsula Freeway, which will be the major bearer of longer distance traffic from the southern part of the Peninsula towards Melbourne, suffers from the same topographical problems that beset the Peninsula’s railways. While the Nepean Hwy does its best to ascend Mt Martha and skirt the ridge of Mt Eliza, this road will now become only be a secondary road.

To avoid these hills, the future Mornington Peninsula Freeway (Frankston Bypass) is projected to continue up the Safety Beach “saddle” and along the flat section towards Moorooduc, then bypass Moorooduc Village to the east, climb towards Langwarrin next to the Stony Point line and descend slowly down to the connection with the Eastlink/Frankston Freeway at Carrum Downs.

Frankston Bypass

By deviating the freeway so far east (even further east than the current Moorooduc Hwy) we are presented with an interesting situation – that a road corridor designed to serve the western Peninsula intersects with a rail corridor that really only serves the eastern Peninsula. The point of intersection will be an overpass over the Mornington Railway at East Moorooduc.

This project would have obvious spin-off benefits to the Mornington tourist railway. For one, the current traffic level on the Moorooduc Highway would fall considerably, improving the chances of reopening the level crossing at this location. Secondly, the road contractors would be required as a matter of law to reinstate the railway after digging it up at the overpass site. And the practicalities of deliberately reinstating rotten sleepers and mature trees into the trackbed would suggest that MRPS might be given the free gift of a restored section of track, at least as long as that disturbed by the overpass contractor.

And of course, improved road access to the Peninsula would boost the number of tourists able to ride their railway.

But what if there was ’something more?’ Something that could be used not just by tourists.

Park and Ride

Many public transport theorists shy away from the idea of Park and Ride for good reason: it is only ever a partial solution at best, and wastes other opportunities for public transport development.

Land near stations is, or should be, valuable for transit-oriented development and to give it away free to commuters is wasteful. Park and Ride undermines potentially viable bus services, at their most viable point. It supports peak flows only, which for most workers are only 2 trips a day, versus a more flexible public transport system which can support more trips per day than 2. Motor vehicle movements near stations create hazard and delay for pedestrians and bus users.

Park and Ride also requires the customer to own a car, and therefore loads into their budget the substantial fixed costs of owning a car, and therefore encourage the customer to minimise the average total cost by driving the car, rather than using public transport where prices for the customer are usually set as a proportion of average, rather than marginal, costs.

But Park and Ride does have some natural territory. Particularly in the urban/rural fringe where bus services of a suitable standard for turn-up-and-go usage will never be provided. Many people live in what is called ‘acreage’, sometimes as a lifestyle choice (having horses, for example, or just liking the country feeling) and sometimes out of economic necessity, as people move to fringe villages like Tynong North or Coldstream that have never had adequate public transport.

Many of these people continue to work in the urban area, and like all commuters, would welcome a choice of options and to not have to face unnecessarily high fuel and parking costs.

While getting a critical mass of such people for a rail service out of such low density population will be a challenge, one saving grace is that the net would be cast fairly wide, and the catchment could be as much as 10-15 km or more from the railhead.

What park and ride currently takes place?

The size and heavy usage of the Frankston station carpark is a pointer to the need for better public transport serving the Peninsula as a whole. With people on the western Peninsula effectively unserved, and people on the eastern side poorly served (but at least served) by an infrequent railcar shuttle, it is little wonder many drive to Frankston.

The current service does not support these people. While a sixty-minute trip, stopping at most stations (or in some cases, all stations, even in peak times) is probably adequate for someone who lives in Frankston and was until recently competitive with the peak road times to the CBD, for someone from further south, the arrangement is unsatisfactory. The access via local roads is also unsatisfactory, with each twist and turn, and multiple sets of traffic lights a disadvantage versus the “keep going” option with driving.

What are the factors in good Park and Ride design?

Existing Park and Ride proposals and actions have generally been poorly considered, and often to justify political rather than transport ends. The first of these is to reuse existing stations such as Garfield or Ballan, in the centre of the respective townships, one senses to justify the station rather than provide genuine Park and Ride options.

The Park and Ride customer, like all customers, has a measure of true-end-to-end journey time. This journey time includes a wait time, transit times on all modes, connection time, and average lateness or delay. The Park and Ride customer is generally a commuter (with 2 journeys a day, to and from work) and would use their car for other journeys, usually in their area.

The fact that they are travelling only to and from work somewhat limits their flexibility of when to travel, but the imposition of standard working hours in most industries then limits their flexibility of when they can start and finish. Which makes the job easier for PT providers, as they therefore do not need to provide a turn-up-and-go service, but provide a timetabled service, which, if close enough to the times for standard working hours, the customer needs to fit their travel timings around.

But common to the Park and Ride journey and the full journey by car is the first section of the journey, by car, usually via a local road and then a freeway, to the point of exit from the Freeway to the Park and Ride. Because at this point the customer has two choices: to diverge to the Park and Ride, or keeping going on the freeway. Every minute spent between the freeway exit and the Park and Ride is a minute of travel where the customer could have ‘kept going’ along the freeway and are that much closer to their destination.

This provides one of the first key lessons of Park and Ride: that building the stations right at freeway overpasses (usually cheap, degraded pieces of land) is the best option, and every stop sign, traffic light or queue that the motorist faces to get from that freeway exit to the train or bus is a nail in the coffin of the Park and Ride option.

There are several sites I have identified around Melbourne where cheap, degraded land associated with freeway overpasses and the acreage belt coincide with the rail network. All would be suitable for Park and Ride and in a couple of cases, this is already occurring. East Pakenham, at the literal end of the wire would be one such place, pulling in the acreage dwellers of Tynong North and Nar Nar Goon North, who might be better able to access the network at this point than at Tynong or Nar Nar Goon stations.

Another would be the former General Motors station site, where the South Gippsland Freeway crosses. This would require new on and off ramps but the land below would be quite suitable for a large carpark and a station on the Pakenham line, which could pull those people in from the very low density suburbs around Hampton Park.

The new Mornington Peninsula Freeway, as noted in the map above, crosses the line at East Moorooduc and again at Langwarrin. The next point of intersection is Kananook, which is from the rail network’s perspective an excellent site for Park and Ride, but from the road perspective a poor one: the freeway easement is very tight at this point. The land around Kananook would also have real commercial value and disturbance to neighbours would be high.

The Langwarrin site has the advantage of being on an existing rail service albeit non-electrified. However, this site, too, will be fairly tight as the freeway descends into a cutting or tunnel under the Pines Reserve and Golf Course. It is also a urbanised area with the same problems with neighbours as Kananook.

Which leaves the East Moorooduc site, which has the disadvantage of requiring maybe 2 kilometres of rail track restoration, and the provision of an additional service not covered by the Stony Point line. The site is otherwise a greenfields site, but which will be degraded by the construction of the overpass, with few neighbours and few other urban land uses.

Viability of a Park and Ride service

As was stated above, the commuter from the southern Peninsula who is bound for inner Melbourne, and wishes to avoid the high petrol and parking costs of travel to inner Melbourne, is the target for this service. There will be no walk-up traffic, and any bus service to this site would need to be additional to the coastal urban service which currently runs. It is possible a few bus journeys from Rosebud and beyond could be scheduled to the Park and Ride to connect with the service, and save considerable time for genuine commuters without cars to make the entire journey.

The time for the commuter is a ‘given’ time from their southern Peninsula origin point to the exit of the Mornington Peninsula Freeway at East Moorooduc, at which point the clock starts ticking as to whether they should keep going, say to the Frankston Railway carpark, or complete their journey by car.

I timed it yesterday, Sunday, in average off-peak traffic conditions, and it is exactly ten minutes from the Frankston-Flinders Rd level crossing on the Mornington line to Frankston Station at Young St. In peak times it would be considerably worse. The new road will not help reduce journey times into Frankston Station, as the road goes north east to Langwarrin then north west to Carrum Downs.

Rail, on the other hand, takes currently eleven minutes from Baxter, a time which I’m sure could be reduced by one minute simply by improving the entrance arrangements to Frankston. Assuming the new line segment only adds one minute to the journey, we have a time which is competitive with road over the same distance.

Such a service would of course cost money, both recurrent and capital, to get it going.

What other budgets might be offset against this cost?

First and foremost, the provision of additional parking at Frankston Station. A few decks of carpark might cost $10 million; a cheap restoration of the 2km from Baxter to East Moorooduc might be similar. A self-propelled railcar would need no run-around. Beyond that point it is all recurrent expenditure.

Operating the rail service could be expensive, but a few substitutions from other possible scenarios would be instructive. For example, a frequently mooted possibility is to increase the frequency of Stony Point services. With no passing capacity south of Frankston (except for the train shunting onto the Long Island Branch) you would need to reinstate one of the old passing sidings, for example, Somerville or Hastings. Crib Point has intact sidings, but is too far south to be useful as a loop.

Again, this project would give you the opportunity to operate a second service south of Frankston, with movements onto and off the short branch taking place while the Stony Point service is south of Baxter (which is most of the time). Safeworking requirements would be minimal.

The service, and the Stony Point one, would become the interurban express service that I have mooted in other places. Frankston commuters would be offered a seat on this train with few stops before the city – at a premium. Leawarra and Baxter commuters would find this a significant advantage as well.

A six-car Velocity or Sprinter set, with say half the seats full by Frankston, and the balance taken by Frankston commuters, should be able to reach the CBD within 60 minutes from Moorooduc East. This includes 11 minutes from Moorooduc East to Frankston, a one-minute stop, then 48 minutes (see 1967 timetable) from Frankston to Flinders Street. This train could be a ‘balancing run’ against a Geelong service that would otherwise have to find a home, such as the one that runs to Caulfield sidings each morning.

End-to-end journey times for the commuter from Dromana to Young and Jacksons (we can’t really know which CBD destination the commuter is going to) should be approximately 75 minutes: 10 minutes to the Moorooduc East park and ride, 5 minutes to park the car and board the train, 60 minutes transit time to Flinders Street, and 5 minutes to exit the station complex. This compares with end-to-end driving time of 75 minutes in a future peak driving scenario (as follows: maps.google.com quote of 1:15 but factoring peak delays of 10 minutes and parking at end of journey, minus saving from Frankston bypass to be built of 10 minutes).

Cost savings would be more substantial – a Zone 1 and 2 daily at $10.10 versus approximately $5 return on Eastlink, approximately $5 return on Citylink, EITHER 8 litres of petrol for a modestly efficient car for the return journey ($24) which is the lower bound with the fixed costs worn by the driver, OR a fully-costed kilometre rate for petrol, wear and tear and depreciation of 70 cents a kilometer, or a massive $117 dollars per day. Even the lower bound (variable cost only) has the driver paying $34 a day plus parking (which using the best early bird rate is still about $12 a day).

In conclusion, I have tried to argue the case for a piece of public transport infrastructure that probably won’t be built. While I think on its own you could argue for it, as part of a package of improvements it would be a long…way down the list of priorities.

Comments?