Personal reflections on urbanism, urban life and sustainable urban design in Wellington, New Zealand.

Wednesday, July 26, 2006

Back on track: choosing rail

This might seem like a strange time to be optimistic about rail, given that fares are about to go up and the Overlander has been axed. The former was inevitable, and while the latter is sad, our sparse population and rugged topography make inter-city rail impractical. But it's metropolitan commuting that's the big issue, which is why I'm cautiously heartened by the Phase 1 Consultation Report (1.4MB PDF) from the recent Ngauranga to Airport study. It summarises the submissions thus:
There is good support for public transport, especially light rail, [my emphasis] with a route through the city to the airport a popular topic. Improved access to the airport, walking and cycling and protection of heritage and urban form were also issues frequently commented on.

It can be concluded that those that gave feedback seem to think that problems on the corridor are not caused by the roads, but the number of cars on them which can be decreased by improvements to public transport.
Of course, that's just what the submitters say, and it's no guarantee that they will be taken seriously. But given that nearly a third of submitters (there were 46 submission in the end) mentioned light rail without prompting, we would be justified in being angry if it it were not included as an option for the next phase of consultation.

While it may superficially sound like a separate issue, the future of the Johnsonville line is vitally relevant, since light rail is one of the options for that route, and the CBD section could then form the first stage of a city to airport link. While submissions on the Johnsonville corridor closed two weeks ago, there's still plenty of public debate going on. While one commentator warned against removing a "popular and affordable train network", John Rusk denies that the Johnsonville line is popular. He bases this upon 2001 census figures showing that "63% of northern suburbs commuters choos[e] bus and only 36% choos[e] rail", and upon observations that at Johnsonville station "there is a steady stream of people boarding a steady stream of buses - to travel from one railway station to another". This seems to run counter to international experience, and certainly would come as news to the Johnsonville line commuters I know who dread the thought of losing trains.

So why the difference? It all comes down to the way you define "the northern suburbs". While the options in the study were all about the future of the Johnsonville railway line, the study area extended to the far-flung outskirts of suburbs such as Newlands and Churton Park. If you include residents more than 3km from the nearest station, it's no wonder that they "choose" the bus! If one were cynical, one might almost conclude that the study was designed specifically to make rail look bad...

But let's not get into conspiracy theories, and instead see if we can see whether commuters "choose" bus or rail when both are available. I took 2001 census data for all meshblocks with at least 10 public transport users, and mapped the proportion of them who used bus (blue) or rail (red). Yellow dots show stations, and the green line encloses areas served by both bus and rail (click on the map for a larger version).

Map of public transport choices in Wellington's northern suburbsThere's still a lot of blue within the joint bus/rail area, but most of that is in parts of Ngaio and Khandallah that are separated from the stations by a deep gorge. A better comparison is Johnsonville, where commuters really do have a choice between bus and train. And most choose the train.

All of which is consistent with US studies, which conclude that when "service conditions are equal, it is evident that rail transit is likely to attract from 34% to 43% more riders than will equivalent bus service". And the remainder who choose the number 56 or 57 buses over the train? Perhaps they're the ones who work further along the Golden Mile, since the buses travel to Courtenay Place, unlike the railway line. Unless it's converted to light rail, of course.


At 7:35 pm, July 26, 2006, Anonymous Anonymous said...

I doubt anybody would choose the 56 over a train: it runs through Newlands for half an hour before it heads down Ngauranga Gorge. It's the 54 citybound (from Churton Park) that takes most of the bus passengers from Johnsonville station, and there is generally a line of around twenty people to get on to the busses between 9 & 10am.

I suspect a large part of the reason people would choose bus over train there is that it's significantly faster - the bus will arrive more than ten minutes sooner than a train leaving at the same time, simply because of the route the train takes (of course, that is different when the traffic is heavy). Most of the passengers stay on further along the route than the Interchange though. I can't see a busway making much difference to that either. The comparison Rusk makes is false in this light.

Including Churton Park (where I live) and Newlands in the survey area was bizarre, since most of either is much to far from Johnsonville for a train to be reasonable. I live in the closest park of Churton Park proper, and it's a twenty minute walk to Johnsonville station. Most of the suburb would be double that, and Paparangi and Newlands even more.

If the train extended out this far (as gets mooted from time to time) it would definitely be more attractive, but I don't think that's on the agenda and it would be very difficult to effect regardless. I suspect the same is true of Newlands-Paparangi-Grenada. It's unfortunately not very likely to occur, although an extension into Grenada via Newlands could be justifiable, and even practical if the line were changed into light rail (enabling it to share the current bridge over the motorway).

At 9:42 am, July 27, 2006, Anonymous Anonymous said...

I dont have much to add other than to say thanks, Tom, for the excellent work. I thought the figures from the study sounded fishy, and the map you've created illustrates why very well.

At 8:01 pm, July 28, 2006, Blogger John Rusk said...


Thanks for the interesting comments. It's great to see so many people getting involved in this debate (even if we don't always agree :-)

>I suspect a large part of the reason people would choose bus over train there is that it's significantly faster - the bus will arrive more than ten minutes sooner than a train leaving at the same time,

That's a very significant point. On the busway, express buses will be able to pass other buses that are stopped at stations. That will go a long way towards adressing the travel time issue - i.e. take passengers quickly, from J'Ville, down what is currently the rail corridor. Similar overtaking is not really possible with rail.

>I can't see a busway making much difference to that either. The comparison Rusk makes is false in this light.

Can you elaborate on what you mean here?

>Including Churton Park (where I live) and Newlands in the survey area was bizarre, since most of either is much to far from Johnsonville for a train to be reasonable.

Exactly! It is indeed bizarre to include those suburbs in any discussion of rail. But it is not at all bizarre to to include them in a discussion of public transport. Upgrading rail will make no difference to you (and me, and everyone else who lives in those suburbs). Only the busway will give us better public transport - and it will do so without degrading the public transport offered to existing rail users. (They still catch public transport from the same stop, and travel the same route. The difference is that its a bus instead of a train, that it's more frequent, and that it runs into the CBD.)

Here's what I find bizarre:
residents of Churton Park and Newlands arguing in favour of rail, when only the busway will get them out of the congestion on the gorge. (Don't take this comment personally, you're certainly not the only one. My point is that the busway benefits you; rail does not.)

At 9:28 pm, July 28, 2006, Anonymous Anonymous said...

I never experience congestion on the Gorge (I never get a bus arriving at Wellington before 9.10), so it is possible that that is a part of it. It is, however, an immediate turn-off for a unidirectional busway, since I'll often be getting a bus against the "peak" flow on the line (I'm a student).

I can't see why passing stopped buses would ever be necessary on a one-way route when all the buses are going to the same place citybound. The other way is a possibility, but I doubt there are actually enough people getting a bus from Ngaio to Newlands for it to be a factor.

The passing that has an effect with the trains is when one has to wait for the other headed in the opposite direction; it's true that this wouldn't happen with the busway, but only because service is degraded versus rail now.

Actually, something occurs to me just now that doesn't seem to be covered in the survey sent around: what happens at the peak direction-change? Surely the entire line would have to be flushed clean, meaning no buses at all for a substantial period. There is possibly some way to deal with that, but I don't see it mentioned here.

I don't feel that your "But it is not at all bizarre to to include them in a discussion of public transport." is actually a reasonable statement. Churton Park, Newlands, Grenada, Paparangi, et al, are all covered adequately by the present buses, both ways, all day. Ngaio (to give an example) is not - the bus service to Ngaio runs only a few times a day. Because, of course, of the trains — but it's only going to be covered by the busway for half the day (in each direction), meaning there'll need to be substantially more buses running (in which case, what's the point?) or a large degradation of service.

People travelling between Johnsonville and Wellington would quite possibly receive a benefit from a busway, but nobody else on the Johnsonville line would as they would with a continuous light rail system, or even with the current trains. So yes, I, in Churton Park, would advocate to keep, or upgrade the rail system because I want to have as many people using public transport as possible, in total, as opposed to a possible saving of a couple of minutes a day myself.

I'm actually skeptical of that saving, since the train route is much less direct than that which the buses take now, but I will grant that it's at least plausible and won't discount it completely. The major problems with the theory are that the buses can't run much faster, if even as fast, as the trains do, and that the train line is a suboptimal route anyway. A bus running along that line Johnsonville-Wellington in thirteen minutes is just implausible.

There are also the many bus stops in between the two along the Hutt Road etc which would need to be eliminated, also diminishing utility. The alternative here is to run more bus services, alternately on the busway and by road, but that's genuinely impractical: Churton Park just doesn't have enough of a population to justify it, and neither does Newlands. If that's done without increasing the total number of buses then again, what's the point of it all? So I can't see any real benefit to me anyway. The busway certainly doesn't give me anything substantial.

If the rail or light rail line ran out this far, I would use it, because I find a train much more comfortable to be in (even on the Johnsonville line), but otherwise I'm happy with the present bus service.

At 2:10 pm, July 29, 2006, Blogger John Rusk said...


You highlight a lot of key issues, mostly to do with the unidirectional nature of the busway. The council's planners need to clarify what will be happening in regard to those issues (there are lots of different options - some are good and some are not) and they need to communicate their ideas clearly.

Yes, outside peak times the existing bus service is OK. Higher off-peak frequency would be nice (and it might be possible too, with the busway) but apart from that the services are OK. At the times you travel, road congestion seldom affects the buses. But, if you were to routinely travel between 9:15 and 8:45, the congestion effects on travel time would be much more noticable(and that's when most commuters do travel).

Passing stopped buses is necessary to mix "express" and "all stations" services on the same route.

You wrote that it's implausible that a bus could run along the J'Ville line route in 13 minutes. I thought so too, until a few days ago. There was an accident on the Hutt Road, and my bus was diverted up the Ngaio Gorge. From the bottom of the Ngaio Gorge to J'Ville station took 12 minutes and 17 seconds. If they can do that on the winding road (with traffic), then they can certainly do it on the busway (with smoother curves and no traffic). Add a few minutes for the portion of the journey which I didn't time (Wgtn station to bottom of Ngaio Gorge) and you get times entirely consistent with the 15 to 17 minutes predicted by the council - maybe even a bit less.

(By the way, why did you pick 13 minutes? The trains do depart at 13 minute intervals, but they actually take 21 minutes to travel the length of the line. In fact, while 21 mins is the scheduled duration, it appears the current trains may take a couple of minutes longer.)

At 11:40 pm, July 29, 2006, Anonymous Anonymous said...

I don't think higher off-peak frequency would be any more possible with or without the busway, there really aren't enough people to sustain it around the Churton Park leg, they're rarely full then anyway. So while it would be nice, it's not likely - geography works against it, as the more frequent bus routes tend to be the ones hitting several different places, which doesn't work well in what's essentially an enclave.

I would experience more congestion travelling at peak times, the Gorge being the major bottleneck there, so bypassing that could be faster: I just don't think it's worthwhile for the few affected services a day. I'm not sure the benefit would be huge regardless, except on the days with the absolute heaviest traffic.

Over a third of the train journey is between Crofton Downs and Wellington, which isn't included on your timing there. That'd be more than half of the current bus journey time already. Even if the buses ran appreciably faster, I can't see it being faster.

Thirteen minutes is how long it takes (in practice) a bus from Johnsonville to the Interchange, which is the most comparable time.

The major objections I have to the busway plan are:
1) It's a service downgrade for current rail users between Raroa and Crofton Downs.
2) Unidirectionality is a substantial service degradation. There'd need to be many more buses running to/from the intermediary areas during the other times... so what's the point, then? Having to use two transport media on the same route is very definitely suboptimal for all sorts of reasons.
3) I don't like replacing electric trains with diesel buses. Some back-of-the-envelope calculations tell me that it'd be at least triple the buses running along that spine just to cover the trains. That's almost definitely an underestimate.
4) I think real rider numbers would drop on a bus-based system.
5) Cost: I actually did an Economics assignment on almost exactly this topic at high school, which didn't work out favourably. The council may well have better ideas of things now, but the long run cost per passenger was actually (very slightly) higher with a busway, taking in to account council survey responses to predict ridership. This was a while ago, and the information it was based on was a few years old at the time (from the last time the Council considered this, I think), so it is possible things have changed (and of course I'm not an expert on the subject regardless), but I'm not convinced there would be any cost saving at all.
6) I like the light rail concept as a general thrust of strategy, and would like to see it implemented in Wellington. If it isn't done this time on this route, I prefer not to rip out the rails there now (separately to preferring to keep rail for other reasons).

I don't oppose the concept of a busway itself: just this one, this time. Putting one in on a new route, or a route currently served with buses, could be worthwhile. My preference is still for light rail in general though, but Wellington's geography works against that sometimes.

At 9:46 am, July 30, 2006, Blogger John Rusk said...


I have to correct two of the facts you've quoted:

You wrote: "I just don't think it's worthwhile for the few affected services a day"

It's not a "few affected services". Take route 54 for a example. 45% of the morning services on that route (that's 45% of all services before midday) leave J'Ville between 7:14 and 8:24 - which is exactly when congestion affects services, as you'd know if you travelled at that time.

Further more, a very high percentage of Northern Suburbs travel is on-peak (over 80% I understand), we're not just talking abouot 45% of services, we're talking about 75% of the bus users, if not more.

You wrote: "Over a third of the train journey is between Crofton Downs and Wellington, which isn't included on your timing there"

My timing was from the bottom of the Ngaio Gorge.

I'd also like to respond to your list of 6 points. I'll bookmark this page an add it to my list of things to blog about sometime.

Bye for now,


At 11:19 am, July 30, 2006, Anonymous Anonymous said...

I still consider it a few affected services in comparison to the effect on present rail users and on the buses for the rest of the day.

The single largest problem to me is the one-way nature of it - I see that as a killer. The rest is just about which system is better, but a one-way busway just doesn't make sense, and a two-way one probably isn't possible.

The other thing that occurred to me is Raroa station: it serves the intermediate and Onslow College (and relatively little else), in both cases against the "peak" flow, for obvious reasons. It'd take at least nine regular (non-busway) buses to cover that. This type of situation is what leads me to think unidirectionality is unreasonable.

I look forward to seeing your post.


Post a Comment

<< Home