The Eye of the Fish

February 8, 2008

Light Rail Redux

In my last post, I expressed concerns with the light-rail proposal as detailed in the Ngauranga-to-Airport (N2A) strategic study. Implicit in my statement at the end of the post that “I have a hard time getting behind this light rail proposal at this point in time” is the fact that a different time or a different proposal could indeed change my mind.

As for better proposals (or lack thereof), the N2A technical report effectively takes them off the table when it notes (on page 16) the following:

A high quality passenger transport connection between the Johnsonville growth node and the CBD has been investigated in a separate study [The North-Wellington Public Transport Study]. A range of options were investigated including retaining heavy rail, replacing heavy rail with a light rail system, replacing the railway with a guided bus system and using buses on existing road network. Part way through this separate study, Council decided that the existing heavy rail network would remain. Consequently, we have assumed that the high quality passenger transport service connecting the CBD and the Johnsonville growth node will be heavy rail.

Given this, my initial inclination was to ignore these and similar options in my analysis of the N2A report. Shortly after posting, however, I was referred to the Transport 2000+ initiative, which rightly notes in their own submission on the N2A study the absurdity of a study labeling itself as “strategic” refusing to lay out a broad strategy due to the current political reality.

While I can imagine Machiavellian maneuverings and political pressures that would lead to this outcome, I believe I am largely immune to these threats. Hence, this post, which was originally intended to discuss options within the framework laid out by the report, will instead focus specifically on options that were explicitly placed outside of that framework. I will, though, be watching my back.

The crux of the Transit 2000+ (T2k) proposal, as was the crux of my previous post, is the topic of transfers. While I argued against the introduction of more transfers into Wellington’s transit system, T2k more proactively argues to largely eliminate an existing transfer: that at the railway station.

Their proposal does not entirely eliminate the need for these transfers, but it does so for many riders, particularly,for shorter-distance rail riders for whom the overhead of a transfer would constitute a larger percentage of their trip time (and therefore a greater barrier to ridership).

The proposal comprises the following: run light rail south from the railway station, eventually to the airport; but also extend the light rail to the north along existing rail infrastructure. In other words, cover the core of the Wellington region with a single, integrated light-rail network. Beauteous!

Importantly, the T2k proposal revealed new possibilities to me by noting that it’s possible to run light rail and heavy rail together on the same tracks. Previously I’d thought that the only Johnsonville could be reached by light rail, by replacing the existing heavy rail line. The ability to run light rail to Lower Hutt and Porirua lets many more people get to the heart of the CBD without requiring a transfer, and opens up possibilities of future light rail corridors into the hearts of those cities, with direct connections to central Wellington. (The T2k study elegantly proposes to maintain existing heavy-rail transit for further suburbs such as Upper Hutt and Paraparaumu, at distances for which heavy rail is more appropriate, and floats the idea of these trains running ‘express’ through the areas served by light rail, providing further benefits to riders.)

What does this proposal do to the ridership numbers south of the railway station, which concerned me so in my previous post? I don’t have any hard numbers, but I can’t help but imagine that it increases demand in the CBD significantly, with large numbers of people who currently terminate their trip at the railway station continuing onto Lambton Quay or Willis Street. While this effect would likely have minimal impact on ridership south of, say Courtenay Place, other factors come into play there. Firstly, a better-integrated transit system is more likely to attract new riders, increasing ridership in general. Secondly, viewing the Courtenay Place-to-Newtown (or -to-Airport) segment as a portion of much-longer routes changes the calculus significantly: the overhead of running the line is amortized over a longer route, and lesser-travelled sections on the ends can be better justified, as it doesn’t cost that much more to run the trains a bit further.

And as for the forced transfers to light rail at Newtown, et cetera, that were naughty in mine eyes before? In the worst case, I see the negatives of these transfers as being outweighed by the positives created elsewhere. Possibly, though, the improved economics behind the rail line make it more feasible to continue running a subset of the current buses into the CBD in conjunction with rail. Again, I don’t have numbers, but I feel it’s worth further study.

And what’s this talk about rail to the airport, mentioned above? The T2k proposal suggests implementing their light rail solution in three phases:

  1. Johnsonville—Courtenay Place
  2. Lower Hutt—Hospital
  3. Plimmerton—Airport

This phased approach makes a lot of sense, with each phase extending further south and adding a new route in the north. However, the N2A report predicts a vanishingly-small ridership on a rail line from the airport, and therefore only really considers implementing light rail as far south as Newtown. Their analysis, though, seems to only address existing transit riders who would choose to switch modes to light rail at the airport. Sitting in traffic waiting to get through the Mt. Vic tunnel the other morning after shuttling visitors off to the airport, I couldn’t help but imagine that a well-implemented light-rail transit node east of the tunnel (with, dare I suggest, park-and-ride facilities) would attract a good number of CBD-bound commuters who’d rather not deal with the hassle of driving every day.

That topic—along with the rest of the eastern-suburbs transport situation—is fodder for a future post. But first, look for some specifics on the nature of the central transit corridor proposed by the N2A study.

As for this post, what have I accomplished? Well, I’ve appropriated the light-rail vision of Transit 2000+ wholesale. But that’s because it’s a good one, and those folks have been thinking about it much longer and harder than I have; I couldn’t hope to compete. Plus, while the group is still active, their website appears moribund, and I couldn’t find their detailed proposal on the web (my source material was forwarded to me by a helpful reader). It definitely deserves to be publicized more widely.

This is the second post in a series on the Ngauranga-to-Airport corridor strategic transport study report, comments on which are due February 22, 2008.

8 - 02 - 08

Good stuff – Am working on a submission with similar recommendations for the N2A study (for what it’s worth – as you point out, they are outside the options posited in that study).
Anyway, you have brought to my attention some useful ‘evidence’, which will be helpful in asserting my case – Thanks!

8 - 02 - 08

There are problems combining light and heavy rail on the same line. Mainly to do with one coming off much worse than the other in a collision. But also with platform heights and widths when different size trains are using the same stations.

Johnsonville is always going to be a problem due to the small tunnels. I’d imagine if you turned this in to a light rail line, then you could extend it to wherever you like and there would be no transfer problem for those Johnsonville commuters. In fact you’d reduce some transfers from the Johnsonville line on to buses at the railway station. Some people would use light rail in the city, especially if they live near a stop or are traveling within the CBD. Others would prefer to continue to use buses. The buses don’t have to go away.

light « Poneke’s Weblog
9 - 02 - 08

[…] the Johnsonville trains with buses) that the Johnsonville line will always remain heavy rail. Eye of the Fish notes “the absurdity of a study labelling itself as ‘strategic’ refusing to lay out a broad […]

9 - 02 - 08

Meanwhile they’re going ahead and spending $5 million on lowering the tunnel floors on the Johnsonville line. Given LRVs would have vastly different dimensions, shouldn’t this been put on hold and perhaps reallocated to other track work.

Just to sound out a few thoughts:

Regarding the idea of Light Rail further north. This is something I’ve heard Brent Efford and T2000 promote before. It is something I agree with myself, but not right now (i.e. in about 10-15 years). Right now, start with jsut the Johnsonville line. The points to consider are:

1. We’ve already purchased new heavy rail vehicles, so too late now. (Of course this wasn’t the case a couple of years ago when this idea was going around.)
2. However at the time (and still for the most part currently) LRVs were not as capable of high speeds, or as much comfort at high speed running as Heavy Rail, so it could’ve been a serious compromise.
3. That is changing as the technology develops, there are some good options that will become available to do this in the future as Tram-Trains are becoming more popular. It would be safe to assume in the next 5-10 years from now good quality, comfortable, high speed tram-train options would be readily available that will no longer be a compromise over their heavy rail cousins.

So lets look at the strategy:

First, plan/start building LRT from Johnsonville to Airport.

Second, continue running with the new Heavy Rail trains we bought until demand increases to such a level as we need more trains. At that point purchase tram-train vehicles to increase capacity by operating into the city for for the shorter routes (e.g. Plimmerton/Lower Hutt).

I would recommend maintaining a separate ROW around the waterfront for these tram-train vehicles through to Courtenay Place (or the intersection of Coutenay Place/Taranaki St) so as to offer faster service. (It may not get used, but if the ROW is reserved in planning, it’s better to have the option.)

Also, I don’t buy this argument about people not using it from the Airport. Why build LRT out to Miramar but not serve Miramar itself? There are obviously a couple of routes that could do this, one of them I’ve shown in my ABC submission (see website). This would put stops in reach of a lot of other commuters. Remember people will tend to walk a lot further for LRT, say 500m, and that doesn’t include people that would choose to ride a bike to the local stop. Or drive to the local stop if you had park and ride. Or be fed by local shuttle services to capture the rest (e.g. similar to the way the 30 and 28 shuttles feed the 11 route).

In terms of some other routes and the argument of having to transfer:

1: The 1 could remain it is sufficiently patroned to do so
2: The 10 would dissapear entirely (assuming the route is similar to what I put in my aBC proposal)
3: The 22/23’s would be questionable, it might be sensible to turn one or both into shuttle services to get people up/down the hills to the Zoo stop.
4: The 3 would be rerouted to go north through Kilbernie and up through Hataitai and the bus tunnel, adding extra services to Hataitai residents that wouldn’t be served by the LRT route.
5: The 91 flyer would no longer need to exist obviously
6: The 11 should remain, but perhaps also heading through the bus tunnel instead of going through Newtown.
7: The 44/43 might pay to become shuttles also feeding into the Miramar stop.
8: The 2 already goes via the bus tunnel.

Those from the eastern suburbs wanting to get to Newtown would take LRT, or would transfer to LRT at Kilbernie before their busses head north to the bus tunnel.

9 - 02 - 08

Thanks erentz, emphasizing a long-term phased plan is a perspective that I don’t think has been given enough emphasis in considering our options (both now, and in the past).

Having that additional line around the waterfront for the trains is an interesting proposal, jervois quay certainly wouldnt miss the lane. The plans with light rail running down the golden mile seem that they would have to compromise speed for saftey – does anyone have any case studies of similar situations?

Tim Jones
9 - 02 - 08

There’s a public meeting on Wednesday 13 Feb, from 6-8pm, at St John’s (cnr Willis and Dixon Streets) to discuss Ngauranga to Airport. For all the details, see:

9 - 02 - 08

I’ve added the public meeting to our ‘upcoming events’ section; thanks.

Flat White
10 - 02 - 08

I don’t think that existing bus routes can be dismissed as easily as erentz suggests. Surely all current bus routes that service areas off the main arterial routes need to remain in some form? For example, I can’t see residents of Houghton Bay walking to Newtown to catch the light rail service to the CBD.

10 - 02 - 08

At the end of my post I floated some possibilities for some of the routes that serve the south/east suburbs. In the case of Houghton Bay it is questionable whether it makes sense to continue the 22/23 routes as they are. The reason is that these routes don’t just serve the people at the very end of the route, they add capacity to the entire spine through Newtown, Mt Cook/Adelaide, and the CBD. Given these would be well served by LRT, the options for example for the 22/23 routes might be:

– Merge them in some fashion.
– Reduce frequency.
– Keep frequency but convert to a smaller shuttle service that feeds the Light Rail stop by the Zoo.
– Maybe a mix of the above.

Houghton Bay could be one of the most affected, as it is the hardest to justify a full and frequent service right into the city for so few passengers.

10 - 02 - 08

regarding possible routes for a modern Light Rail system in Wellington: surely the best place to start would be to look back at the original tram routes in Wellington, before they ripped them out and replaced them with buses and trolleys? There would be a clearly established heirachy of routes and settlement patterns established on that…

11 - 02 - 08

I see the main route for light rail as running from Johnsonville to the Airport via the Golden Mile, Adelaide Rd and Newtown. Later, it would make sense to convert the heavily trafficked former tram route to Island Bay back to light rail (which uses articulated tramcars anyway).

But it would not make economic sense to convert other bus routes back to trams.

I would retain all the trolley bus routes (with the later exception of Island Bay) through the CBD, running them via the same Golden Mile transit mall as light rail would follow, but stop most other bus routes at the CBD fringe or at light rail interchanges… at Newtown for the Houghton Bay, which as has been said, hardly warrants a bus into the CBD for the passengers carried.

However, I would convert the other end of the 22/23 (the Mairangi route) to trolley buses, as this would enable the heavily used 17 and 18 routes which serve the university also to be converted to trolley buses, given that most of the wires the 17 and 18 would need are already in place.

Running all those Hutt, Eastbourne and Newlands buses through the CBD causes chronic congestion. Before 1991, they terminated at the Railway station. Once light rail was in place, I would terminate them there again.

The only non-trolley exception to buses through the CBD would be the 14, but that used to be a trolley route, sigh. The traffic offering doesn’t really warrant re-stringing the wires, and I doubt wires would be allowed along Oriental Bay now.

11 - 02 - 08

Running a light rail on the existing Johnsonville line seems so logical, it wouldn’t share any of the line with heavy rail until the station approach. I would prefer it to run from the station along the quays (there are 6 lanes to play with) and then up Kent/Cambridge Tce (there are 8 lanes to play with) to Newtown & eventually in a later stage to the Airport/Miramar. The golden mile would still be clogged with buses & popular for short rides so the speed of LRT will be to it’s advantage on it’s own dedicated ROW along the wide avenues.

I however do see a possible short term problem with the line being off the golden mile, it may not attract the short trips through the CBD. But the airport connection and it’s clear ROW would be the making of it.

This would also have the added advantage of removing a lane of traffic from our potentially beautiful waterfront ‘avenue’.

11 - 02 - 08

Just to clarify for some. I seem to talk about the affect on just the southern/eastern ends of existing bus routes a bit, but not the western and northern suburbs. That’s because to a very large degree those parts of the existing routes would not require any change.

Alex: I used to think there was too much traffic in the CBD for LRT to run efficiently through the Golden Mile. But the research indicates it’s really the best place for it, LRT works best when highly visible and servicing pedestrianised places. In terms of speed you have to take into account:

– A dedicated public transport ROW through the Golden Mile
– Intersection priority over vehicles
– LRVs carry more passengers per vehicle, meaning there are less busses in the city (an often over looked one that has a big impact, e.g. one typical LRV carries 4 busloads of people)
– At key choke points there is the option for separation between the bus routes and the LRT routes by sending busses one way, and LRT the other, e.g. between Taranaki Street and Willis Street

12 - 02 - 08

In his enthusiasm for Light Rail Poneke says Running all those Hutt, Eastbourne and Newlands buses through the CBD causes chronic congestion. Before 1991, they terminated at the Railway station. Once light rail was in place, I would terminate them there again..

What about the fact that “The benefits of extending public transport services from the railway station through the CBD are highlighted by the large patronage increases achieved when Newlands bus services were reorganised in 2000 and extended from Lambton Bus Interchange through to Courtney Place (that is, a seamless service was implemented through the CBD). This resulted in a 40% increase in patronage as frequency increased and a seamless service provided where passengers were able to travel through the CBD without having to interchange.” -North Wellington Public Transport Study Stage 2 Report.

Of course terminating the bus services at the rail station will undo the above and force 40% of 2,700 bus commuters from Newlands, etc. back in their cars. And, of course we will still be expected to pay “our share” of Light Rail to end up with (for us) a poorer public transport service.

Hey, where’s an idea . . . why not fix the CBD so Wellington can have a real Bus Rapid Transit. When we get a public transport service that is better for all.

12 - 02 - 08

Running a light rail on the existing Johnsonville line seems so logical”.

This has been looked at many times (1984, 1991) and always found wanting. Most recently the 2006 North Wellington Public Transport Study (NWPTS) found, Light Rail to J’ville would reduce car users by only -453 (-2.5%) and increase PT usage by +377 (+10%) when converting the line to a busway was nearly as effective (–348 and +296). It also found the 25-year Light Rail Cost (Capital plus operating) (Johnsonville-Courtney Place) to be a staggering $385M . . . three time the current heavy rail option and $100M more than the busway equivalent.

Overall the NWPTS estimated the Benefit to Cost Ratio for the Light Rail option was by far the worst at 0.08 (i.e. costs are more than 10 time the benefits !). Even with Light Rail to Johnsonville half the PT users would still go by bus !!!

Any when you think about it, it does make sense. North Wellington is very hilly and low density (the two are related). So access to the rail stations is difficult for a majority of us, as indicated by the census showing the highest area of bus usage is Ngaio . . . East ! This, in turn, means that, even with a rail line, North Wellington has lower PT usage than either South or East Wellington and running Light Rail out here will not make much difference.

12 - 02 - 08

Following my earlier posts, here is a comment from someone who thinks Light Rail is a total waste of money. Just look a the costs and benefits.

Just look at the cost per seat. A seat on a standard bus costs about $5,000, a trolley or hybrid $10,000 and a fuel cell bus $20,000 and a Light Rail vehicle about $35,000 per seat. Of course the Light Rail cost is just the vehicle, if you “include installation” of $116M, then the roughly 700 seats from 8 Light Rail vehicles costs a staggering $200,000 per seat !!. Forget Light Rail costing more than a car, this proposal has each Light Rail seat costing nearly as much as a bus !!!!

And then lets look at the benefit. This is modelled and the impact of car usage is outlined in the technical report. Just compare the “Base Case” (table E1.4) figures with the Light Rail (table F13). This shows Light Rail will reduce car users by a massive -278 (-0.2% of car usage) and increase PT usage by a staggering +155 (+0.5%) !! As this is with the model already set-up to have users prefer rail over buses. In other words, the modeling predicts the $140M Light Rail option will make almost no difference to Wellington traffic.

Forget Light Rail, it costs too much and does to little.

12 - 02 - 08

Tony, “Of course terminating the bus services at the rail station will undo the above and force 40% of 2,700 bus commuters from Newlands, etc. back in their cars.”

It is pretty hard to judge whether that would be the case without serious thought. Times have changed, fuel prices are up, congestion is up, LRT is known to attract passengers from a much wider area, etc. But on the face of it for the northern suburbs it probably makes sense to keep the existing services runing through to Courtenay Plc. It certainly would be a mistake I think to from day one decide to reshape the entire network just because LRT is in. You’d want to do it over a couple of years as it settled in, and you could see how it was actually affecting patterns.

Also people seem to be stuck on the idea that this is about now. It isn’t, it is about planning for the 30 year horizon and where the populations are going to be at that point. This is something the reports get seriously wrong. They all are based on 2001 out to 2016. First 2001 is already too out of date to be used in an report that decides $650 million dollars of spending. Second, 2016 is not strategic in any sense.

One of the amazing flaws in the report was its assessment on population growth, which is a measly 0.65% per annum for Wellington City. Actual population growth has been double that at 1.27% per annum since 1991, and nearly three times that at 1.84% over the past 5 years.

Then there is the fact that the growth in the areas along the spine has been even greater, and is planned to accommodate the majority of Wellington Cities future growth over the next 30 years.

Doesn’t take a rocket scientist to work out that there is a serious problem there that affects the viability of everything in the report.

“A seat on a standard bus costs about $5,000, a trolley or hybrid $10,000 and a fuel cell bus $20,000 and a Light Rail vehicle about $35,000 per seat…”

You cannot use that kind of mathematics. It it like saying 1000 metres of 2-lane road can accomodate 250 parked cars. It just doesn’t mean anything. What matters is the cost for the throughput of the system you put in place.

And yes, LRT may be more expensive on the face of it, but one would have to do a full cost analysis of the bus solutions (including environmental cost, cost of roadway space (why is it assumed to be free for a bus but not LRT), cost of catenary for trolley busses, etc. After that you also need to look at the difference in benefits, the vision for your urban environment, your growth plans, etc.

“This is modelled and the impact of car usage is outlined in the technical report.”

The Technical Report is absolute bananas on so many levels. I can’t imagine writing something like that in my line of work and using it as a basis for a $6.5 million investment, let alone contemplate it for a massive $650 million investment and the related economic consequences which are even greater! We (Wellington) really should be asking for another study done by another outfit.

It worries me that the politicians aren’t saying the same thing, it leads me to believe they haven’t properly read the reports, or don’t understand them and so are falling back to their ideolical viewpoints on the subject.

12 - 02 - 08

Undoubtedly LRT is a hefty investment, but the cost analysis process seems like an unbelievably complex process, and something that defies simple comparisons.

Thanks for the link, Rapid bus is very interesting.

13 - 02 - 08

If the problem is congestion then the solution is not building LRT. The problem is that roadspace is not marginally priced. If it was then there would be far less congestion, buses would move more freely and be cost competitive against cars, and freight would also win. The tired old solution (tired because it was tried in the US in the 1970s and failed) of building extremely expensive single use infrastructure (rail) wont fix congestion.

LRT is an expensive busway with the disadvantage of two levels of bespoke infrastructure (overhead and track), whereas a busway is essentially the current road with some lining, signing and enforcement. Low emission buses are substantially cheaper than LRT and don’t forget Wellington is getting “new” trolley buses at a substantial price (with upgraded infrastructure). Why duplicate that again? This is about allocating precious resources (and I like trams), and paying for Rolls Royce public transport when there are other needs is just wasteful. Most people who would use LRT would ride buses. Similarly without road pricing, most people who drive will continue to do so.

Rail technology is good for high volume movements of people at high frequency, which simply doesn’t exist in Wellington. The truth is that any major investment in public transport will lie mostly idle most of the time underused, but still costing money that can’t be recovered from the users. The problem is that peak demand isn’t priced, which if it was would spread demand over a wider period and across modes better.

13 - 02 - 08


I’m not sure if you’re proposing congestion charges or other ways of making drivers realize the actual costs of driving (by removing the current subsidies), but I agree that such measures are a much more holistic solution to traffic (and other) issues.

I do have to question these claimed failures of rail in the US. I’m not sure about the 1970s, but there have been a number of very successful (light) rail projects in the US in the ’90s-’00s, that are drawing greater-than-expected ridership, especially impressive given the sprawling nature of many of the cities in question. One of the most successful systems has been in Los Angeles, of all places, along with Portland and even places like Dallas and Sacramento, which are extremely car-oriented.

I do still have some qualms about whether Wellington has the numbers to really justify light rail (as you say, poorly-spent transit funds will come back to bite you in many ways), but I feel it’s better-justified when viewed as a long-term strategic move. There are options that were not sufficiently addressed (or addressed at all) by the N2A study, and they deserve to be.

Also, I think you trivialize the difficulty (and the cost) of doing BRT well, which is the topic of my next post.

13 - 02 - 08

Libertyscott, “Similarly without road pricing, most people who drive will continue to do so.”

Quite true. The report (as far as you can use it for anything) concludes more or less this. By 2016, with all the improvments, the same congestion will be experienced. So given one of the key issues is free and easy movement of freight and commercial traffic, it would seem to be a waste of time building more roads. Better to start with congestion charging first to discourage private use, and free the roads up for that commercial traffic.

Kevyn Miller
15 - 02 - 08

If peak oil has the sort of drastic impact that many are predicting where will the money for public transport come from? Increased funding for PT has added over 4cents a litre to the price of petrol since 2000 compared with an extra 5 cents for roads.

Are the cost estimates current ones? The 40% increase in the construction price index since 2000 doesn’t only affect roads.

21 - 02 - 08

Also, I think you trivialize the difficulty (and the cost) of doing BRT well, which is the topic of my next post.

Still looking forward to your post on href=””>Bus Rapid Transit . . . it is a pity it will not be before submissions close on the Ngauranga to Airport Strategic Study.

Eye of the Fish » Honey pot
18 - 02 - 09

[…] Light Rail proposals in Wellington […]

Togs, Togs, Undies | Eye of the Fish
19 - 12 - 09

[…] Light Rail proposals in Wellington […]

4 - 01 - 10

Thanks erentz, emphasizing a long-term phased plan is a perspective that I don't think has been given enough emphasis in considering our options (both now, and in the past).

Having that additional line around the waterfront for the trains is an interesting proposal, jervois quay certainly wouldnt miss the lane. The plans with light rail running down the golden mile seem that they would have to compromise speed for saftey – does anyone have any case studies of similar situations?