Monday, June 14, 2010

The June 2010 Caulfield group train timetable - service trends

One significance of the latest train timetables is that it is the first to have commenced after the arrival of new trains. This has allowed additional peak services to be scheduled, in this case largely on the Caulfield group.

To assess the significance of this change as against others I went through my old timetables and counted the number of trains that passed Caulfield in the morning and afternoon peak periods. Suburban trains that arrived at Flinders Street between 7:00 and 8:59am or departed between 4:00 and 6:59pm were counted. The results are below:

Timetable date: No trains (am) / No trains (pm)

20 Jan 1975 (VR): 42 / 47

15 Apr 1985 (Met): 31 / 39

29 Aug 1988 (Met): 35 / 43

21 Oct 1991 (Met): 34 / 42

6 Dec 1998 (Bayside): 28 / 38

19 Nov 2000 (Bayside): 29 / 40

27 Jan 2002 (M>Train): 30 / 40

Oct 2004 (Connex): 30 / 40

15 Oct 2006 (Connex): 32 / 42

30 Sep 2007 (Connex): 33 / 43

27 Apr 2008 (Connex): 34 / 45

Nov 2008 (Connex): 35 / 46

20 Jul 2009 (Connex): 37 / 46

June 2010 (Metro) : 42 / 55

There will have been other timetables not featured above. A weakness of this simple count method is that it neglects cases where services have been extended from middle-suburban locations such as Mordialloc and Oakleigh to outer suburban termini such as Frankston and Cranbourne. So although train throughput may have fallen, service kilometres may have held up due to average longer runs.

However some broad trends are discernible, such as the following:

* A decline in peak service around 1980 (this was the point that train patronage reached historic lows), a partial recovery later that decade until another drop in the early 1990s.

* The 2000s saw a weak recovery, adding a train or so per year, though it wasn't until 2008 that train numbers had been restored to 1988 levels.

* 1975 peak service levels were only met or exceeded in the June 2010 timetable. Off-peak service levels were boosted in the 1990s, followed by Sunday service a few years later. At most stations evening services remain less today than in 1975 (every 30 instead of 20 minutes) due to service cuts in 1978.

* The effect of new infrastructure such as the third track to Moorabbin. The timetable released just after that (1988) included extensive express running that was pruned in subsequent timetables until being increased again in 2010.

* The growth of outer areas on the Dandenong line such as Pakenham and Cranbourne. Until this latest timetable this meant that nearly all timetable additions on the Caulfield group went to those areas rather than Frankston.

Comparison with patronage

How do the trends in service level compare with patronage? Patronage underwent a long decline from the late 1940s to 1980, with the drop in the 1970s the steepest (reaching a trough of 93 million trips). This fall partially reversed in the 1980s, with strong rises for several years. However the state's recession, high cancellations (due to industrial disputation) and service cuts in the early 1990s caused the decline to resume.

Later passenger numbers levelled off and started to grow slightly. For example, Department of Transport evidence provided to the Legislative Council Select Committee on Train Services indicate little patronage growth between 2000 and 2004. At best passenger numbers rose in line with population, making no progress towards achieving either the franchise contracts ambitious patronage targets or the modal share aim of the Melbourne 2030 plan.

Comparing this account of patronage with train numbers up to this time indicate a correlation - higher service levels were associated with higher patronage and lower service levels were associated with lower patronage.

2004 marked a turning point where patronage diverged from service level offered. Overcrowding increased as more passengers started riding substantially the same number of trains. At the time the government was grappling with new train contracts after the walkout of National Express. New trains arrived but these merely replaced the scrapped Hitachis rather than expand the fleet. National Express suspended driver training, causing cancellations to rise due to driver shortages. And when Connex took over the whole system their priority was to reintegrate the system by, for example, training drivers who had only been taught 'their' half.

By 2009 annual patronage had exceeded 200 million - more than double that of 1980 and about 50% up on just five years before. Extra trains were slotted in here and there, but with passenger growth exceeding service growth, overcrowding and delays intensified.

Public transport as a political issue rose in profile and the government ordered more trains. This order was later increased. The new trains started arriving this year.

The June 2010 timetable was the first to take advantage of these extra trains. This allowed a peak service increase unseen for many years. Meanwhile patronage is still growing but slower than the steep rises of the 2000s. Hence we are in a period where peak service is growing faster than patronage, marking a catch-up on the 2004 - 2009 period when patronage rose faster.

1 comment:

Riccardo said...

PP I'm not sure why you do these exercises, exonerate the political establishment and ignore obvious implications from what's happened.

Who approved the extension of suburban trains and urban sprawl to Pakenham? It only had electrification by historical accident (a failed briquette haulage scheme) and was not suitable for the level of service that would need to be provided.

If the train kilometres increased, it means they borrowed/stole them from existing commuters on existing services, it's no wonder they fell as a proportion of total ridership. And now people can't get on at Murrumbeena because of unsustainable services at Pakenham.

And why no comment on how reductions in services made the service overall unappealing?

I know you mean well but I always feel I'm reading the Jim Betts version on your blog.