In a kind of parochial way I have always wondered which of our rail lines perform better and it is also something that I have seen others debate at various times. The most simple way to do this would be to just look at AT’s regular patronage reports which gives a breakdown of patronage for the Western line and the group of Southern lines which is made up of the Southern line, Eastern line, Onehunga line. The graph below shows that the Southern lines carry quite a few more people than the Western line, in fact they are carrying over 60% more people.
To the naked eye it looks like the Southern lines are growing faster than the Western but when comparing the patronage of each of the lines vs the total number of rail trips we see that very little change and what is there tends to favour the western line. To me one of the big problems with this is that the Western line is kind of all by itself while what we are comparing it to is in effect 2.5 lines, hardly a fair comparison so with this in mind I thought I would do a little investigation to work out an answer once and for all.
First I wanted to deal with the biggest issue which is there are far more services that head South and East. To do this I have gone through the current timetable (the one before Manukau was added) and counted up how many services there are during a normal week. This is shown in the table below:
As you can see only 30% of all services are on the Western line yet comparing it back to the patronage, the line currently carries 38% of the people. Now just comparing the total number of services is useful but it creates it’s own issue in that not all services are created equal. A service from Onehunga to Britomart is obviously shorter than one from Swanson which itself is shorter than one from Pukekohe. As a result I took it a step further breaking down the number of services based on where it starts from and where it terminates. Multiplying these numbers by the distance for each service get a total number of Km’s that trains are travelling while ‘in service’. We don’t get patronage information weekly so I have then multiplied that by 51 weeks to get the distance travelled by trains in a year. Now before anyone tries to point out that there are 52 weeks in a year by reasoning for picking 51 weeks is that throughout the year there are a few days that services don’t run, often these are replaced by rail buses but not always e.g. Christmas.
After all of that we then get the results below which show a very interesting story. It shows that for every in service Km that trains do they carry an average of 3.8 people however when looking at the individual lines the Western line is leaps ahead with 4.7 people per Km vs 3.4 for the group of Southern lines, a massive 38% difference. The Western line has also been growing faster than the lines out South so this difference may keep increasing, Manuaku should help but will it be enough. The other thing to remember is the Western line is also still stuck with only 4 trains per hour during the peak, perhaps it is time for AT to look at reallocating some resources to where they are needed.
After doing this I thought it would be interesting to see how this has changed over time. The current timetables go back to September 2010 and so I have also done same analysis on the timetable prior to that and the results shocked me. I had expected to see a drop with a timetable as it more services will take a while to attract customers and in the case of the western line that took 9 months but as you can see below, on a per Km basis the Southern lines are currently carrying less patronage than they did in August 2010 and what’s more it has shown no signs of improving. We can only hope that Manukau brings in enough patronage to not only offset it’s new services but also to make up some ground lost at with the last change.
Processing...
I think an important factor in these statistics is that the Western Line is used quite a lot for inter-peak travel that doesn’t go to Britomart. People heading to Henderson, New Lynn and Newmarket in particular I suppose. This means the trains are often quite well used throughout the day.
What this highlights is the stupidity of continuing to have such incredibly poor weekend services on the Western Line.
Agreed Peter, it’s hard to believe that Swanson still has the same hourly service it had a decade ago, and still no trains at all on Sundays.
Quite a good metric – one that perhaps could be used in AT board reports. Far better than the cumulative patronage figures they use at the moment.
I’d also be inclined to believe that the type of catchments on the Southern/Eastern Lines aren’t like the ones on the Western. The Western Line features catchments for New Lynn, Newmarket and Henderson business districts (all of which are far more extensive and progressive than the ones in Glen Innes, Ellerslie, Papatoetoe and Manurewa), Unitec at Mt Albert, Eden Park at Kingsland, and then there’s the distribution of schools along the line. The most that the Southern/Eastern Lines can boast is Sylvia Park, Mt Smart @ Penrose, and the odd airport connection which I doubt is used as a transfer from rail… The Southern and Eastern Lines position themselves well for daily commuter peak traffic, but as for “anchor destinations”, there are few – I’d believe Sylvia Park is the largest anchor attraction on this side of the network, aside from Britomart and Newmarket… I don’t think the Southern/Eastern/Onehunga Lines are in any trouble from slow growth – this will shoot up when Manukau is linked to the South, when EMUs arrive to improve reliability, and finally when the CRL is built and opens up the possibility of improved frequencies.
You’re right, in that the Southern and Eastern Line is still very much commuter orientated – there’s been little attempt to integrate things. Otahuhu is an example of things being completely wrong. Glen Innes is an excellent example of things being half right. There’s a large university campus (UofA’s Tamaki Campus, 300m from a train line. Yet there has been no attempt to connect the two, and there is no plan to do so in the the next few years. I’ve talked to the authorities and even a pedestrian crossing is off the agenda (I’ve seen near fatal misses). So much missed opportunity.
In fact, it’s about to get worse, as AMETI will cut a de-facto motorway through the community, and split the university from the train line completely. We claim to be thinking about public transport and communities, but there’s still a lot of narrowly focused limited-use thinking going on.
I think that it shows how important it is for us to develop destinations along the Southern and Eastern lines to take advantage of the infrastructure that exists.
I honestly don’t expect substantial patronage increases on the Southern and Eastern while the trains remain so unreliable. Delays and unreliability form a natural constraint on anything but long-distance peak-hour commuter travel, which is similarly beset by delay. I should mention that in much of Auckland peak hour seems to be spreading between 4pm and 6.30, which is disturbing but another argument for (more) reliable rail capacity.
If oil prices sustain and the dollar slides, or oil prices increase and the dollar sustains/slides, then we’ll see another push-driven increase. If not, we’ll see natural growth, and some uptake from Manukau. I’ll have to take a look and see that line soon.
Station improvements are also making a difference, I’m sure. I’ll be interested to see the daily usage figures per individual station when they’re collated again this year.
I would point out that trains out west have been just as unreliable as those out south
If not more so, but the unreliability is on a single line not three of them. Onehunga manages spectacular levels of reliability, but it’s doing it on four trains an hour and they’re only two cars so the risk of terminal mechanical faults is lower than a six-car-plus-locomotive Western Line train. The Southern and Eastern have the same trains as the West, but they mostly travel longer distances and they’re running services with an aggregated higher frequency than the Western Line.
The Southern/Eastern lines won’t see patronage equal to their service kilometres until there’s no period during which you could wait at a station for 47 minutes before 9pm (see my post below) on a week night without seeing a single train. The nature of the lines militates against people just turning up to stations, whereas the Western Line is nearly offering service frequencies where “forget the timetable” is a viable travel option.
What caught my eye from the first graph is the dramatic rise in passenger growth around the Rugby World Cup (expected), but also that it didn’t drop back straight afterwards.
I’d like to see AT give out free day-pass tickets out in the CBD carparks to encourage people to *try* public transport – it might be enough for them to break their habit and see that PT is possible. Still lots of people in Auckland haven’t caught a bus in their lives.
Like this?
http://www.thelocal.de/national/20120410-41851.html
Yes Tim and Pub. But not yet in my view. See George above, the product just isn’t good enough at the moment. I would love to be marketing PT in Ak but it would be a mistake to be pushing it too hard until it is better; can actually deliver. Probably take the operation of the EMUs and the bus network redesign to be in place, oh and actual integrated fares, then it will be a wild pleasure to be selling it.
Right now we should focus on selling its potential…. See the previous post…
We love your work AC + AT; but how you are communicating your plans needs improving, anytime you want some constructive criticism…?
…..*sitting by phone*
I’m struggling with the “passengers per in service kilometer” concept. Surely this is a meaningless measure unless you know how far those passengers traveled? For instance, a single train running a single 10km journey with a single passenger onboard leads to a measure of 0.1. Where as if the same service stops every kilometer and the single passenger gets off and another gets on the train then they’ll have carried 10 passengers for a measure of 1.0. So what? Why would we want to measure this?
Surely passenger kilometers traveled is much more useful, and maybe dividing that by the service kilometers would be useful in measuring passenger density on a particular line?
A good point Obi. Not sure how easy it is to find that information though.
Perhaps Matt’s analysis shows us that Western Line trips are shorter than Southern/Eastern/Onehunga line trips. I think that’s a fair assumption.
I think 10 people getting on and going one stop are more valuable than one person going ten stops. All that parking saved (so long as they don’t Park ‘n’ Ride), for a start.
Perth have a documented plan to increase PT useage through events services. If the experience of those who use events services occasionally is positive, they are more likely to become regulars. So it’s really important to get the events services working well.
Sydney and Melbourne have weekend special fares. In Sydney family tickets are $2.50 per person (Sundays only) and cover unlimited travel on trains, buses and ferries. A service I took last Sunday in Sydney had 40 passengers on one top deck, plus about 10 on the adjacent mezzanine floor. If this was representative it means there were about 800 passengers on the 8-car train. Many of the passengers were family groups from outer suburbs going to Circular Quay and other locations in the city. The train is then a “fun machine” by children in the family groups, which must then become a positive experience affecting their transport choices later in life.
A great promotion for Auckland would be to have similar family tickets available during school holidays once most of the electrification works are finished – perhaps spring next year. There is little point in promoting a service that’s unavailable many weekends because of line works, because this just increases the cost of providing substitute bus services, where the marginal cost of additional passengers is greater than for the train. Greater weekend patronage is likely to increase the use of central city facilities such as the waterfront and the Botanic Gardens.
Obi,
Matt’s number is actually “Journeys per in service km”, not “passengers per in service km” (he can’t know currently where each passenger got on and off the network so counting journeys is the best metric available).
In your example, 1 journey for 10km is a value of .10 (10 km/1 journey), whereas the second is 10/10 = 1.0 (journeys per in service km).
Same metric would come out if train travelled 9 km empty, then 1 km before end 10 people got on and all got off at the 10th km, ratio would be 1.0 there as well.
(or train travelled 9km with 1 only person who got off at the 9km mark, then 9 got on for the remaining 1km)
The problem with metrics like this is that they are averages of averages (averages over a year and averages over all services on a given line both peak and off peak and averages over all journeys taken on the line), so they don’t/can’t tell you anything about the peaks, how big they are, when they occur, and where in the network they occur.
Still you have to start somewhere with numbers.
When the HOP system comes on board properly, then the tag on/off data will allow real time numbers to be calculated and used.
In the London Underground system they have over 7 million live “HOP” equivalents (Oyster cards) and they know where and when each and every card was/is being used in real time (they do use that real time data to catch fare evaders/scammer amongst other things).
Of course, not everyone will use HOP, but even if a majority do it will be a major step forward in gathering such metrics on PT use when it does in real time basis if needed.
A big difference is also the nature of the Western vs the Eastern/Southern/Onehunga. The former is a single line, the latter is three lines with separations and convergences. As a passenger who lives north of Otahuhu, I can only use the Southern Line or the Onehunga Line, I cannot use the Eastern Line. For passengers north-bound from south of Otahuhu, if they want to get to a station between Otahuhu and Britomart they cannot use Eastern Line services. The divergence of trains that travel most of their service kilometres on the same set of tracks is not a convenience. It means that some of the densest-packed settlement areas along the lines only get, if they’re lucky, a train every 20-30 minutes (there are gaps of 19 and 20 minutes through Ellerslie in the morning peak! And the evening has a train at 19:41 and then the next isn’t until 20:28) – after Otahuhu there’s precious little density around many of the train stations – whereas until you get to the outer extremes of the Western Line you do have every train available to you, and you know that it’s going to be 15, 20, 30 or 60 minutes until the next train based on service frequencies. The other lines have irregular frequencies because of the divergence, and we know that irregular waiting times are as bad for patronage as lengthy ones.
Does the low patronage in mid 09 – 2010 have something do to with the double-tracking that was happening at that time?
I would think so. Train reliability plummeted around then from memory.