Wellington Scoop
Network

Coming soon: bustastrophe 2

bus-direct

by Glen Smith
Till the recent network changes, Wellington had a bus system that consisted almost entirely of ‘direct’ bus routes. People travelled on a bus which continued all the way without transfer to the CBD (or beyond in many cases) with many buses sharing a common route for part of the journey. This design served Wellington well for decades.

The main central common route was the Golden Mile which had reached capacity and was unable to accommodate growth. One option to solve this was to keep direct services but increase capacity by adding a second public transport (PT) corridor. Until recently planners refused to add this second corridor and never presented the option of continuing all major routes as ‘direct’ services.

Instead the bus network changes incorporate a Golden Mile based ‘trunk and feeder’ model for many routes, which forces a large number of commuters to transfer to larger central units at peripheral ‘hubs’ to complete their journey. This reduces the number of ‘units per hour’ along the common corridor but imposes a transfer penalty that discourages the use of public transport. This change has been one of the major causes (in addition to several others) of the ‘bustastrophe.’

bus-feeders

You would have thought LGWM would learn from this and reconsider the ‘trunk and feeder’ design. Instead they plan to increase it and impose this model on to the majority of future PT users across Wellington.

Our rail service already fits a ‘trunk and feeder’ model for across-town riders from the north, who have to transfer from the four rail lines on to a bus-based across town ‘trunk’ at the Station – (compare with the right hand side of the ‘trunk and feeder’ diagram above). The effects of this can be seen in the fact that although ‘through’ traffic makes up over half of car transport trips from the north, only 15% of rail users transfer to this bus trunk – all the others are in their cars swamping our congested roads. The objective should be to fix this basic flaw. But instead the planners want to impose the ‘trunk and feeder’ model on bus users as well – and a very large number of them. As Kerry Wood noted in discussion in a recent wellington.scoop article (8 August) –

“MRCagney have proposed that all bus routes south of the MT Vic tunnel and east of Adelaide Road and Island Bay Parade should terminate at a MRT hub: Miramar, Kilbirnie or the Hospital.”

This ‘trunk and feeder’ model is apparently not open to public discussion or review by our elected representatives but, like the Basin Flyover, has been planned as a ‘fait accompli’. We are told (by the same architects of the recent bus network changes) that it will result in a superior PT design with improved transport times and greater PT uptake. But does this claim stand up to critical examination? And is this the design that the people of Wellington want?

The topic of ‘direct’ vs ‘trunk and feeder’ design is summarised in the excellent planning guide by ITDP. The discussion is based on BRT but whether the ‘trunk’ services are ‘steel’ (rail) or ‘rubber’ (bus) based makes no difference to the analysis. In their summary they note that

“The conditions under which the conversion of direct services to trunk-and-feeder services will bring overall benefits are fairly limited. Until recently, trunk-and-feeder systems were proliferating in conditions inappropriate to their use, and direct service options were being neglected.” 

The authors note that the choice of ‘direct’ or ‘trunk and feeder’ design is dependent on the setting. The key factors are the number of routes being aggregated, the volumes on these ‘feeders’, and the length of the ‘trunk’ section relative to the ‘feeder’ sections. However other important factors are whether demand is peaked or flat and whether imposition of the ‘trunk’ causes lines to deviate from the optimal ‘direct’ route. Trunk and feeder is ideal where

“.. there are a lot of small direct service routes that overlap for a long distance with the trunk corridor…[so]… if the small direct routes along the trunk route are turned into trunk-and-feeder services, customers on each low-demand route would share the use of a much larger vehicle”

A good example of this is feeder buses at Waikanae or other Kapiti towns (which behave much like the outskirts of a much larger city) where the trunk distance is very long, or small Hutt hill suburbs where the volume in each suburb is low.

On the other hand

“If the number of routes is low, the demand will be split among a smaller number of routes, so the original routes will already be using larger buses. Further, if the routes share only a short section of the trunk, the distance available for larger vehicles to reap benefits would be smaller, so there would be less benefit.”

A good example of where not to impose a trunk and feeder model would be routes 21 and 22 (long feeder line with short Golden Mile ‘trunk’) yet this is what was imposed. In summary

“The greater the number of routes being combined into a trunk route, and the greater the proportion of the total trip that is taken up by the trunk portion of the trip, the greater the benefit. “

bus-potentials

This ‘Theoretical Maximum Benefit diagram’ shows that ‘trunk and feeder’ design has to have a large number of feeder routes with a long trunk section relative to feeder section before you get even a few percentage of advantage in cost benefit.

bus-graph

The graph doesn’t extend below the ‘zero percent line’ but in a large number of cases ‘trunk and feeder’ design will be inferior, especially where a small number of lines are aggregated (less than around 6) and the ‘feeder’ line is equal or greater length than the ‘trunk’.

Our current rail design for riders from the north of the CBD travelling to the southern CBD or beyond is the worst possible scenario (large feeder length and volumes, with a short trunk south of the station). Changing the mode of this across town ‘trunk’ to LRT or trackless trams won’t change the deficiencies of this flawed basic model. This can only be achieved by removing the ‘feeder and trunk’ transfer at the Station.

The analysis notes that ‘trunk and feeder’ design generally requires a larger operating fleet at higher capital expense

“Fleet requirements: When demand is peaked, as is generally the case, more fleet is required for trunk-and-feeder services than for direct services. When demand is flat, fleet requirements are basically equivalent … There will be some routes that will be pulled off their original route… This causes additional delay for all customers and additional operating time and cost for the vehicle operator.”

This is what is proposed for commuters from the eastern suburbs and airport, not only due to detour via Newtown for all eastern PT users, but via Miramar for airport commuters and those in the southern Miramar Peninsula such as Seatoun and Strathmore.

The calculations in this article only look only at financial cost/ benefit rather than whether patrons will actually use a service. In particular it appears to only include the ‘actual’ transfer penalty rather than the potent ‘pure’ (‘disincentive’) transfer penalty. However they note that

“In general, most customers would prefer to take a bus directly from where their trip begins to where it ends … Avoiding transfer delays is often one of the main reasons discretionary riders elect not to use a system.”

“As a result of this transfer penalty, an average trunk-and-feeder system starts with a disadvantage of twenty-three minutes of customer travel time per route converted. These additional delays also translate into increased cycle time (TC), which directly translates into increased fleet needs and increased operating costs. When added to the extra fleet needed to compensate for the peak hour correction factor and extra fleet required at the terminal for scheduling adjustments in real-world conditions, it is typical for trunk feeder systems to require from 11 percent to 66 percent additional fleet. In most cases, these extra costs will total something in the range of 34 percent of an additional cost for the trunk-and-feeder service, while the benefits from the larger vehicle use are going to be generally in a similar range only when the trunk is very long relative to the feeder route, and there are many very small routes converging.”

Some trunk and feeder advocates claim this design gives a more regular dependable service but the authors note that

“while some claim to prefer trunk-and-feeder systems to reduce irregularity of service inside the….trunk corridor, there is little empirical evidence to suggest any significant overall system-wide benefit of this. Having irregular arrivals of feeder buses at transfer terminals where conditions can become severely overcrowded….is as problematic as having irregular direct services with a dispersed impact along the…. trunk corridor.”

In a city the size of Wellington, where most of the population lives within 7 or 8 kilometres of the CBD, a ‘trunk and feeder’ design is a poor choice. The alternative is to continue to use ‘direct’ services for all major services via a ‘radial connective’ design (continuous across town lines from one peripheral location to another peripheral location). This basic design is used by cities much larger than Wellington (see the London underground and Moscow’s rail network for two examples).

To achieve this, additional Quays capacity could be utilised to add additional ‘direct’ across town lines for
all major regional routes. These could be divided between bus-based Golden Mile lines and rail-based Quays lines giving the potential to remove the potent ‘trunk and feeder’ rail to bus transfer penalty at the station to produce seamless regional rail lines (by ‘tracksharing’ of ‘lighter’ units on our existing network) while allowing almost all bus riders to continue to have direct services to the Golden Mile. See diagram below for one possible route design.

bus-routes

When professionals offer advice to clients it is normal to provide a range of options. A real estate agent shows prospective buyers a range of houses, a bike shop offers a wide selection of cycles, a health professional a range of treatment options. The public of Wellington should have the right to decide they want to retain the basic network design of ‘direct’ transport lines across our city (a design that has stood the test of time and is more logical in a city the size of Wellington with its relatively short total line lengths) rather than having to suffer a fractured ‘trunk and feeder’ which unnecessarily imposes delays and transfer penalties on a high proportion of travellers at ‘hubs’ only a short distance from the CBD. This would be the case even if a ‘direct’ network was more expensive (customers are commonly prepared to pay more if an option suits their needs more) but the evidence above is that a ‘direct’ network would be cheaper.

Unfortunately the flyover debacle and ‘bustastrophe’ have shown us that the organisations responsible for our transport planning (the NZTA, the GWRC and the WCC) appear incapable of this basic professional approach and instead follow the philosophy of making secret unilateral decisions behind closed doors, ignoring public feedback and imposing fixed predetermined agendas.

So look out for bustastrophe 2, coming soon to a city near you and brought to you by the same people who gave you bustastrophe 1 but on a grander scale than the original.

36 comments:

  1. David Mackenzie, 20. August 2019, 11:04

    Brilliant. Excellent solution. Therefore it can’t happen!

     
  2. Mike Mellor, 20. August 2019, 11:39

    Interesting analysis, but a few notes of caution.

    1. The analysis relies heavily on the ITDP planning guide, but they are not a neutral source. As they say “our area of expertise is bus rapid transit (BRT)”, and their guide is naturally weighted towards what BRT can do well, one of which is providing a one-seat ride, without transfers. But other modes have balancing advantages, such as the smoothness, relative predictability and permanence, capacity and ubiquity of rail-based transit. “Whether the ‘trunk’ services are ‘steel’ (rail) or ‘rubber’ (bus) based makes no difference to the analysis” is not in fact the case, confirmed by the number of people in the Wellington region who have been transferring between bus and train for decades.

    2. “This ‘trunk and feeder’ model is … not open to public discussion or review by our elected representatives” is not the case – everything in Let’s Get Wellington Moving will have to go through the normal consultation process, including approval by GWRC and WCC.

    3. A good rule of thumb to follow is to avoid being the first or last user of any particular technology. This includes “‘tracksharing’ of ‘lighter’ units on our existing network”, something that has yet to be defined, let alone implemented. (Tram-trains do exist, but have not caught on as a mainstream solution, and Wellington would be a pioneer in some respects.)

    4. Another good rule of thumb is not to over-complicate. A stand-alone mass transit scheme is a complicated beast in itself, and making it part of the wider rail network, developed for different purposes and operating in different ways (and apparently without any input from KiwiRail) makes it a lot more complicated and therefore expensive and prolonged. Adding junctions to the existing system will tend to reduce capacity as will replacing existing trains with lighter ones, and these factors need to be taken into account.

    5. Reading an article like this, emphasising the negative, largely tends to strengthen the impression that transit is all too hard and we should build roads instead.

    6. What appears to be “the best” is often the enemy of the good, and holding out for “the best” at all costs means that we may never achieve the good. Working together on common ground is the key!

     
  3. John Rankin, 20. August 2019, 16:12

    I second @MikeMellor’s comments and add some questions.

    1. How many physical tracks are proposed for the “Quays Rail” section? If it’s 2 (one up, one down), as the LGWM 2 rapid transit lanes proposes, then if I get on at the railway station, every fourth train is going to Miramar, every second train is going to the airport, and every fourth train is going to Lyall Bay. Is that correct? Since not every train from the north will through-run, this appears to imply an infrequent service south of Kilbirnie. Why is a shorter 20 minute rail service to Miramar or Lyall Bay better than a slightly longer 5 minute service, or taking a feeder bus from Lyall Bay to Kilbirnie, as LGWM proposes?

    2. How many buses will be running on the Golden Mile during peak times? The recommended maximum for a reliable service is 60 per hour (one a minute) and the LGWM proposal AFAIK achieves this.

    3. What is the evidence that transfers per se are a major cause of the bustastrophe? As I read the evidence GW has presented, the problem is that transfers are hopelessly unreliable. Have I missed something here? I prefer to suspend judgement until GW has fixed the reliability problems (assuming it can do so).

    4. What does “The public of Wellington should have the right to decide” mean? If the answer is a referendum, that would be saying someone who rides a bus is qualified to evaluate a public transport system in the same way that someone who uses a toaster can design an electricity network. If the answer is a properly constituted and informed citizens’ assembly, then I support this. The professionals from whom the LGWM programme sought advice presented a range of options, with pros and cons of each. They presented these options objectively, avoiding the loaded language used in this article.

    5. Why are London and Moscow relevant to Wellington? These are both underground services, which allow the designers to overcome the physical space constraints of a surface route through the city centre, at significant extra cost. As question 1 above explores, it appears that the proposed “radial connective” design sacrifices frequency to save money and space. What frequency is proposed on the Quays Rail segment?

    6. AFAIK, most cities planning rapid transit put serving hospitals high on their priority list, as hospitals are often the biggest destinations after the city centre. Why does this proposal not see the hospital as a priority for rapid transit service?

    I’d qualify Mike’s comment that “the best is often the enemy of the good” to suggest that we need to work together on common ground to make the good as good as it can be. I’m not convinced that the LGWM proposal is there yet, but that’s what the next stage of planning and consultation is for.

     
  4. Glen Smith, 20. August 2019, 17:07

    Mike.
    1. Can you explain how the trunk being rubber based or rail based makes any difference to the calculations and conclusions. The ‘attractiveness’ (smoothness/permanence/’ubiquity’) of rail, which research shows encourages a greater number of discretionary riders to transfer from their cars than buses, is one of key reasons I support rail (the main reason is the potential to remove the Station transfer). However this is a separate issue to network design.
    2. You say that the trunk and feeder model is open to public discussion and review. Can you point out where alternatives have been presented? You can’t vote or argue for an alternative if one isn’t presented. After Guy Marriage’s article of June 27th I stated to Chris Calvi-Freeman that LGWM was acting in the same way as the NZTA did with the Basin Flyover. He didn’t deny it but stated that
    “LGWM’s recommended programme of investment and the government-funded indicative programme of investment HAVE AGREED UPON (my emphasis) a SINGLE (my emphasis) mass transit spine running from the railway station through the CBD, Te Aro (Taranaki St) and Newtown then via a tunnel to Kilbirnie and through Rongotai to the airport and Miramar.” Not many alternatives there. He then stated
    “The route WILL (my emphasis) serve the CBD through to Taranaki Street…then through to Adelaide Road … The value of stops at the hospital, Newtown central and the zoo are all self-evident. The Aquatic Centre, EBIS and St Catherine’s College are in easy reach of a stop in Kilbirnie and the ASB stadium would be served by another stop nearby. Finally, stops at the Airport and Miramar Central WILL (again my emphasis) be self-evidently popular.” WILL is not a word that implies choice.
    3. Tracksharing is not new or adventurous. http://citytransport.info/Share.htm shows track sharing in a range of countries. http://onlinepubs.trb.org/onlinepubs/tcrp/tcrp_rrd_47.pdf outlines Germany’s extensive experience with this. CCF indicated LGWM has excluded this solution. The reasons he gave were
    a. that trains “able to run on light rail type alignments in the city centre and on the heavy rail alignment north of Wellington station are incompatible with freight trains running on the main lines”. This flies in the face of overseas examples where freight trains do ‘trackshare’ with light units.
    b. That tracksharing “cannot be accommodated within the tightly-spaced ‘paths’ taken by commuter trains at peak hours.” He again offered no qualification of this nor how he proposed to achieve the expansion of rail capacity that has been identified as required north of the Station.
    c. “Power supply” (by which I assume he means different voltages) which is solved by dual voltage trains
    d. “Platform height”. Build new platforms to the same height as our network.
    e. That this was “unaffordable and therefore inappropriate for a city and region the size of Wellington.” I pointed out that “you provide no evidence that the cost of this would be unaffordable… No major construction would be required and changes would be operational. A high-quality train control/signalling system would be required (but this is required anyway) and the ‘pinchpoint’ north of the station addressed (but this requires addressing anyway).” I suggested that if he was certain track sharing was impossible this must be based on a business case and asked him to provide this but none was forthcoming.
    Did LGWM consult with any international specialists from cities where tracksharing operates, before they excluded this option?

    4. “..making..[a mass transit scheme]..part of a wider rail network..” is only one end of the proposed ‘trunk and feeder’ corridor (which could always be tackled at a future date if the Quays corridor is constructed to be compatible). The other end is the single spine winding its way via Newtown and Miramar, forcing almost all bus users to transfer to it, and then take a Quays route. This is poor design compared to adding a number of ‘direct’ rail lines and continuing with ‘direct’ bus lines via the Golden Mile for almost all routes (a couple of lines would have to switch to rail via the Quays to reduce the Golden Mile overcrowding but this should be kept to an absolute minimum).

    5. My article doesn’t ’emphasise the negative’. It objectively compares two options for achieving the same transit.

    6. ‘..”the best” is often the enemy of the good’ implies we shouldn’t even try looking for the best solution. My philosophy is the opposite. Aim for the best and only accept less if the obstacles are insurmountable or overwhelming.

     
  5. Lim Leong, 20. August 2019, 17:36

    With all these LGWM transport experts deliberating on future transport models/options, has anyone asked what the commuters/customers want? The era of solely relying on expert advice on any design is truly over. Now is the era of user-centric design and putting customers’ requirements at the forefront. If you ask a commuter/customer what is the preferred option for public transport, most if not all are going to say “direct point to point”. Is a 100% point to point network achievable within the money and resource constraints? Unlikely, but what efforts have been put into minimising transfers and transfer penalties?

    It has been over 13 months and the much touted transfer-based bus system still is not working. If you are a user, what confidence do you have that the transfers-based system is ever going to work? Another 3 years? Another 8 years? The fact is many commuters are being seriously inconvenienced on a daily basis.

    The overwhelming impression for commuters is that LGWM is debating in the Ivory Tower while Rome burns.

     
  6. Glen Smith, 20. August 2019, 17:36

    John. Run out of time sorry. Will try and answer you after work tomorrow.

     
  7. greenwelly, 20. August 2019, 17:38

    Sigh. Here we go again with “heavy rail” or some variant south of the station. This is a thread that will not die.

     
  8. Glen Smith, 20. August 2019, 19:16

    Lim. Absolutely PT users prefer ‘direct’ service, but eliminating all transfers by a ‘point to point’ system (a network with services going directly from each peripheral destination to every other peripheral destination) isn’t logistically possible. ‘Radial connective’ is the best compromise. Each line goes direct from a single peripheral destination to a ONE other peripheral destination via the CBD (and every point in between). Getting to any destination on any other major line only requires one transfer. It is of course possible to incorporate a ‘trunk and feeder’ model on to these lines (this is effectively what occurs with bus ‘feeders’ on to the Kapiti line). My article is about when it is best to do this and when it’s best to have ‘direct’ lines. For all of Wellington City, ‘direct’ is clearly the best option.

    Greenwelly. The article isn’t about ‘heavy’ rail nor is it mainly about extending our existing lines across the CBD. The main discussion is around the best design for bus services in our city (‘direct’ as we have had for many decades and which worked well, or ‘trunk and feeder’ for which the only attempt has been, unsurprisingly, a disaster).

     
  9. Northland, 20. August 2019, 19:47

    People who ride the buses *are* qualified to evaluate the public transport system they use. They can evaluate:
    1) how late the service was
    2) how unreliable it was
    3) how uncomfortable the ride was
    4) whether they were forced to stand up because no seats were available
    5) how many times they were forced to make a transfer to travel across town
    6) how long it took to make their journey
    All of these evaluations have been made by thousands of Wellingtonians in the last year when comparing today’s bus service to the one from pre-bustastrophe days. We don’t need a citizen’s assembly to work out that today’s service is inferior to the old one.

     
  10. Mike Mellor, 20. August 2019, 21:48

    Glen, I won’t bore Scoop readers with the minutiae of tracksharing: suffice to say that the sharing of ‘lighter’ units with heavy freight trains happens in very few places in the world. None of these have the characteristics of Wellington, so we would be a pioneer, with the issues that Chris Calvi-Freeman raises. Doubtless they could be solved, but at a significant cost. No major construction would be required only if you exclude the junctions with the existing network, which would have to be grade separated; the new platforms that would be required, with their associated trackwork; the additional loops that would probably be needed; signalling alterations; etc etc. It will all add up to an awful lot of money, plus lengthy region-wide service disruption – both much more than for a standalone mass transit system, which would provide most of the benefits.

    And “Coming soon: Bustastrophe 2” and “So look out for bustastrophe 2, coming soon to a city near you and brought to you by the same people who gave you bustastrophe 1 but on a grander scale than the original” read like scaremongering to me!

     
  11. Dave B, 20. August 2019, 22:36

    To all and sundry: Heavy Rail south of the station is what Wellington needs. This was recognized in the 1960’s and nothing has changed to invalidate that. If anything it needs it now more than ever. Whether it can be achieved or not is the question, but it is a question few are even asking. We have become far too hung up on trying to create a new light rail system from scratch, which in most iterations will be of limited use regionally. Yet we have a major problem of regional PT connectivity to solve, and extending the regional rail system that we already have is the obvious answer. Why aren’t we going all-out to try and achieve this? We should only resort to sub-optimal alternatives if despite our best efforts this cannot be pulled-off. But no effort has gone into this since about 1970.

     
  12. Ross Clark, 20. August 2019, 23:42

    To all and sundry: Heavy Rail south of the station is what Wellington needs. This was recognized in the 1960’s and nothing has changed to invalidate that.

    Agreed. The point is that there are two separate problems we are trying to fix here: first, how to connect Wellington better to its CBD – where, in the interim, running express buses down the Quays would take at least some of the pressure off the Golden Mile – and second, how to connect the wider region better to its CBD. A rail extension into the CBD would do this.

    I know much is made of providing through-access (the infamous ‘four lanes to the planes’ or even ‘trains to the plains’), but I don’t think there is enough of a market. The through-market is not a peak or even a journey-to-work market, so the two things needed to get much public transport take-up are not there. The other thing that should be noted, is that the sort of rail network we have means that we cannot support the sort of frequency which would be needed for any through-market to work properly. Once every twenty minutes will not get much off-peak take-up.

     
  13. John Rankin, 21. August 2019, 11:13

    @Northland: I agree that people who ride the buses are very well qualified to describe the problems with the bus network. But as I understand @GlenSmith’s post, he wants the public to choose how best to solve these problems, which is an entirely different matter.

    I remain of the view that as part of the business case for rapid transit, LGWM ought to seek expressions of interest from qualified suppliers, inviting proposals for how to solve Wellington’s problems. That’s the only way I can see proposals like Glen’s and @DaveB’s getting a proper hearing. It would not surprise me if a hybrid option emerged from this process, perhaps extending heavy rail through the city centre plus a lighter solution to the south and east, in future extending north to Kaiwharawhara, Johnsonville and beyond.

    I don’t know the answers, but asking suppliers the right questions and being open to creative solutions would be a good place to start.

     
  14. Dave B, 21. August 2019, 15:59

    @ Ross. You have often commented to the effect that the demand for “through-Wellington” PT is not great enough to justify extension of rail (and being consistent, you also believe the road-demand is not great enough to justify 4-lanes-to-the-Planes). Maybe you are right, but intuitively it doesn’t seem so. Do you have any solid evidence?

    We are talking 100,000 people, or around 20% of the region’s population resident in the area south of the CBD, plus many businesses and regional amenities apart from just the airport. And yet all transport-links are perceived as inadequate, peak traffic congestion is a big issue and public transport is a slow and inadequate bus service. This is very true between Southern/Eastern destinations and the CBD. It is very, very, very true between Southern/Eastern destinations and the rest of the region. That public transport patronage for cross-regional journeys is currently low must be heavily influenced by the service not being there to provide for it. This does not equate to potential demand being low.

    Many times I have made PT journeys across this divide and can attest to its awkwardness. Many times also, I and those I have lived with have either not made such a journey at all, or else have gone by car because it is so fraught. How many do you have to multiply “me” by, before you have a traffic problem, a deleterious dislocation in regional-connectivity and a compelling case for doing something about it?

    Ross, are you possibly being affected by the Auckland Harbour Bridge syndrome? The belief that it was not needed because few people were attempting to cross the harbour before it was built?

     
  15. Glen Smith, 21. August 2019, 20:38

    John. Answering a few of your points
    1.- I had proposed two Quays rail lines- more would be hard to justify. Lyall Bay and Miramar are only possible examples (although with some good justification) and the timetabling/ route/ mode options for ‘direct’ services are again open to modelling based on demand patterns. So for example Lyall Bay could remain as all bus based via Newtown, and Miramar more frequent direct rail (or vice versa). Or bus and rail could alternate for both. Or Miramar could be serviced by ‘direct’ through bus from the Miramar heights between rail services, and Lyall Bay by buses from Hungerford road. etc. The key point is that ‘direct’ services are generally superior to ‘trunk and feeder’ lines. However you seem to want to force everyone on to a single ‘trunk and feeder’ spine based on an assumption that people prefer a more frequent ‘trunk’ service involving transfer to a less frequent ‘direct’ route. Does this hold up to scrutiny? Lets take the 20 ‘direct’ service to Miramar. Most people know when the service is going to leave so get there a few minutes before, but even if they arrive randomly the average wait would be 10 minutes. They then travel by the fastest direct mode without transfer. Going by ‘trunk and feeder’ on the other hand involves an average 2.5 minute wait (assuming your optimistic 5 minute feeder timetable), then the ‘pure’ bus to rail transfer penalty of 17 minutes along with an optimistic ‘actual’ transfer penalty of 3-5 minutes, and then the detour delay via Newtown of an optimistic 3 minutes (good luck with that one) gives 25-27 minutes. Not looking so good. Which is why people hate transfers.

    2. The number of buses along the Golden Mile would be based on modelling and balancing demand between the Golden Mile and Quays over time. However it should be kept as high as reasonably possible since the Golden Mile will remain the predominant destination, and forcing more people (other than ‘through’ commuters) onto a Quays corridor (as your proposal entails) imposes an unnecessary walk. The theoretical optimum is 60 but the Golden Mile has historically functioned quite well at higher volumes. The Ian Wallis review modelled a peak flow of 101-116 per hour.

    3. Transfers are a major problem. Part of this is the unreliability which is impossible to avoid in Wellington where most suburban buses run among mixed traffic with inevitable delays/ variability in arrival times. Direct lines remove this uncertainty/ unreliability.

    4. The public absolutely have the right to choose (via public feedback and their elected representatives- not a referendum) and don’t have to know the exact intricate design details of a public transport network any more than you know the exact chemical composition of a paint you buy or the exact electrical wiring of a flat screen TV. It is the specialist professional’s job to outline what each option offers and costs. Our planners haven’t “presented a range of options, with pros and cons of each.”

    5. London and Moscow are good examples because the larger a city is the more likely a ‘trunk and feeder’ design is likely to be the best (based on trunk length – like the Kapiti example). Yet most large cities still use a basic ‘radial connective’ design with ‘direct’ lines for almost all major lines in their network. Hmmm. Perhaps they have learned something from experience

    6. The Hospital is a priority destination which is why I have placed it on the major north to south and east to west bus lines (where most of the patients are going to come from) and not on the rail line (which most patients, who live on bus lines in Wellington, are going to have to suffer a transfer penalty to reach).

    7. I’m glad you don’t think that ‘the LGWM proposal isn’t there yet’ but don’t quite understand why you think the next stage is going to fix things when it is clear major decisions have already been unilaterally made behind closed doors without public consultation.

     
  16. Glen Smith, 21. August 2019, 21:24

    Mike. Bear with me as I question you on some of the ‘minutiae.’

    What does “none of these have the characteristics of Wellington” mean? In Karlsruhe in Germany the first “line was (and still is) also used by international express passenger & heavy freight trains and all services continued as if nothing special was happening” and now “some services share tracks with all manner of (international) freight and passenger trains – including sometimes the high speed ICE (InterCity Express)”. This is different to Wellington how? Why would the new platforms be any more expensive than the ones you plan to build?

    Why would ‘additional loops’ be required rather than just a simple junction? What is the ‘associated trackwork’ over and above what would be required anyway for the light rail corridor and increasing capacity on our existing network. The stand alone ‘light rail’ proposal I saw needed a huge storage yard with rail and sheds occupying prime waterfront space.

    Why would junctions have to be grade separated? None of our current trains are grade separated and I can show you a design which involves minimal at-grade cross junctions. Why would ‘signal alterations’ cost any more than we require to make our existing system safe and run the light rail spine you propose? Why would this require ‘lengthy region-wide service disruption’ over and above what is required anyway to fix the ‘pinchpoint’ north of the station?

    Why would this “all add up to an awful lot of money”? Please supply your estimates/ costings along with the likely savings from increased PT patronage leading to lower congestion/ pollution/ noise/ accidents so the public can assess if the benefits are worth the cost.

     
  17. Glen Smith, 21. August 2019, 21:39

    Dave B and Ross Clark. When you say you want ‘heavy rail’ across town, what do you mean in practical outcomes? If you mean things like freight then I think the engineering involved is unrealistic. However if the main or only practical outcome is that passenger units can run on our existing network and then across town then I put it to you that ‘track sharing’, as proven overseas, achieves the same outcome without the huge expense. This being the case, why do you feel ‘heavy rail’ is required?
    The best way at assessing ‘through’ traffic potential for any across-town PT spine is to look at Terrace Tunnel traffic (which will be an underestimate) since this is all traffic that starts from north of the main CBD and exits to the south of the main CBD. Based on LGWM data this is around 50% and on WTSM data more like 58%. More than enough to justify a high quality dedicated rail spine.

     
  18. Kerry, 21. August 2019, 22:20

    Glen. You seem to have missed the point about MRCagney’s proposals. Yes (nearly) all southern and eastern suburbs bus routes will terminate at a mass rapid transit connection. Passengers can easily make a change quickly enough to make a faster trip overall. The exception is passengers who will have to change twice: not good.
    Take a look at MRCagney’s Study 13 for LGWM, the last map but one. Transferring passengers have these options:
    — Changing at Miramar, Kilbirnie or the Hospital to either MRT or Route 21, which runs by Courtenay Place and Ghuznee St to VUW, and some buses run on to Karori.
    — Changing at Kilbirnie or the Hospital to either MET or Route 2, which goes by Hataitai and Courtenay Place to the Railway Station and Karori.
    These two extra options make for far greater flexibility that the change-or-walk approach you seem to be thinking of.
    And I agree with John & Mike about the ITDP. They are great for bus rapid transit. But BRT is useless in Wellington. It needs four bus lanes at stops, and no Wellington routes are that wide.

     
  19. Mike Mellor, 21. August 2019, 22:30

    Glen, quick responses.

    1. Karlsruhe is one of the very few places where light rail shares track with heavy freight trains. Hardly anywhere else has followed that example, and none that use narrow gauge. Why not? Because the cost and complexity rule out the option except in very specific cases.
    2. There would either have to be high-level platforms through the city to match the current rail network, or low-level platforms would have to be added to the latter. High-level platforms in the city would be more expensive than low-level ones, and more difficult to fit into the urban environment; low-level ones on the main line would require extra space, often difficult to find, and would probably require gauntlet-type track to stop passing freight trains overhanging passenger platforms. That’s two sets of points per platform per direction, and points are expensive and add complexity.
    3. If the junction between the urban rail section and the current main line is not grade separated, capacity would be reduced by trains having to cross lines in the opposite direction on the flat, and operation would become less reliable because perturbance in one direction would quickly spread to the other. If I were KiwiRail, I would be loath to take that risk.
    4. Rolling stock will need storage, whatever its type.
    5. Standalone light rail requires very little in the way of signalling.
    6. Loops would probably be required because of the different characteristics of rolling stock and of operation between the existing rail line and the new urban section.
    7. With separate main line and light rail systems, any disruption is limited to that system; if they are linked it will spread from one to the other.
    8. Alterations to allow light rail vehicles on the main line will be over and above anything needed to improve the existing system.
    9. Just like you, I have no detailed costings – and it’s a bit strange to ask for it when you’ve produced no figures on costs and benefits yourself!

    Many cities have built and are building standalone light rail systems, with 20-odd such systems opening around the world this year alone. Fully fledged track sharing, as would have to happen here, is very rare, and non-existent on narrow gauge such as ours. Perhaps we should do as you suggested earlier, and learn from the rest of the world?

     
  20. Guy M, 22. August 2019, 4:44

    Glen – “None of our current trains are grade separated” – that’s massively wrong as far as I can tell. I’m thinking of the Kapiti Line, which has grade separated junctions at Kaiwharawhara / Ngauranga, at Takapu Valley, at Kenepuru, at Pukerua, and the place where all the accidents happen is at Paekakariki, where there is no grade separation (but should be). Up the Hutt line, the roads contort themselves to frequently writhe up and over the river and the rail.

    Conflict between train and cars should be designed out as much as possible, as early as possible. Anywhere in Wellington, where conflict between train, car, and pedestrian / skateboarder / scooter will be at a maximum, grade separation should also be at a maximum. That’s why we are talking about either Lambton Quay or the Quays, rather than Featherstone St, which would be much simpler and straighter – because there is less potential for conflict.

     
  21. greenwelly, 22. August 2019, 9:41

    @Guy. The Wellington Metro Network is far from grade separated. Melling is the only one with no at grade crossings (basically because it’s wedged against a motorway for most of its path).
    The Hutt Valley line isn’t too bad, but once you get to Trentham there are a bunch of road level crossings (and there are some pedestrian ones in the lower valley too).
    The Kapiti line has a pile of level crossings (both road and pedestrian) through Tawa, + there is Plimmerton/Paikakariki/Otaihanga/P’Pram/Waikanae.
    Johnsonville has 3 Simla/Khandallah/Raroa.

     
  22. Glen Smith, 22. August 2019, 9:48

    Mike. Let’s probe a little deeper.
    – “none use narrow gauge”- how does the distance between rails make any difference to a network’s function?
    – “the cost and complexity rule out the option except in very specific cases”. What are the ‘costs’ and ‘complexities’ (aren’t they all the other things we were discussing’?) And what are the ‘specific cases’ and how do they differ from Wellington?
    – “There would.. have to be high-level platforms through the city”. Yes. Were you planning a system without platforms? So potentially you have 250 people waiting for a train standing on what?- the sidewalk?- who pedestrians have to somehow get past? and the 250 people exiting the train have to somehow manoeuvre through?. Have you done any modelling on platform size required for the expected passenger numbers?. By my calculations probably at least 4 metres – especially for the main ‘midtown’ stations. Thats why a ‘station’ approach is required rather than the stupid curbside shelters with one or two passengers as shown on the LGWM artist’s impressions of the Quays corridor
    – “high level platforms would be more expensive than low level”. Really??
    – “require gauntlet-type track to stop passing freight trains overhanging passenger platforms”. There would be no freight trains on the across town section. All other stations on the main line would remain the same. Across town units/ platforms would be designed to match current network.
    – “If the junction between the urban rail section and the current main line is not grade separated, capacity would be reduced by trains having to cross lines in the opposite direction on the flat”. Exactly. Lines merging and crossing is one of the major causes of the ‘pinchpoint’ north of the station which could be largely solved by separating the Hutt and Kapiti lines (rather than having them merge) and using existing grade separation points so that exiting of ‘across town’ units occurs separately for each line and without influencing the other.
    – “Rolling stock will need storage, whatever its type.”. My understanding is our rail yard has ample storage.
    -“Standalone light rail requires very little in the way of signalling.”. So would the across town section of a ‘tracksharing’ system.
    -“Loops would probably be required because of the different characteristics of rolling stock and of operation between the existing rail line and the new urban section.” Sorry but what does that mean?
    -“With separate main line and light rail systems, any disruption is limited to that system; if they are linked it will spread from one to the other.” That is true but it is true of any linked rail line.
    -“Alterations to allow light rail vehicles on the main line will be over and above anything needed to improve the existing system”. Really? The only thing I can see is signalling. How much more would a signalling system that allows tracksharing cost compared to the one we need anyway?
    -” Just like you, I have no detailed costings.” This is the point. It is not my job to do this (or yours) anymore than it is my patients’ job to know and present options around the complexity of medications. That is what the planners are supposed to be doing. Perhaps if one of the LGWM team reads this they could present the costings they used to rule out a tracksharing system so the public can assess if they agree with their conclusion.

     
  23. Dave B, 22. August 2019, 13:54

    @ Glen Smith: ” When you say you want ‘heavy rail’ across town, what do you mean in practical outcomes?” The terms “Heavy Rail” and “Light Rail” are misnomers, given that the trains that run on either can be of similar weights. So when I refer to “heavy rail” above, I simply mean something compatible with what we already have in terms of passenger-train capacity. I have not considered that freight might be carried over a rail-extension southwards, at least not in the context of involving train-consists that are any longer or any heavier than the current passenger trains.

    The principal aspect that I advocate is that any rail extension should be able to match the capacity and safety of the lines feeding into it. This implies a maximum throughput of 30 “Matangi-equivalents” per hour, although this capacity would not be used to the max other than for short “bursts” where there might be 4 or 5 trains in succession 2 minutes apart, followed by a catch-up gap. This intensity of operation (with potentially 8-car consists) effectively rules out any form of on-street running or level-crossings. In other words, we need a dedicated rapid-transit corridor free from intrusion by any other traffic. I see it as hugely disadvantageous to opt for a system of trains that are compatible with the public street-environment and then expect the existing rail system to adapt to this. Yes it can be done, but it would compromise the effectiveness of the intensive operation that we currently run. Far better to design any extension to be compatible with this. I see no advantage in departing from a system that demonstrably works so well. What I see is a misguided attempt to do things on-the-cheap – with the real risk of delivering something that is inadequate, probably not cheap, and which cannot readily be superimposed on what we already have.

     
  24. Kerry, 22. August 2019, 14:46

    Glen. The narrow gauge problem is that many modern light rail vehicles are 100% low floor, without bogies as such. They are builtin to the main frame, and designed for standard gauge wheels outside the frames or metre gauge wheels inside the frames. KiwiRail’s 1067mm gauge doesn’t leave enough for a normal wheel-width clear of the frames. Another difficulty is that metre-gauge light rail is seldom the same width as the usual for standard-gauge light rail. 2.65 m wide would normally have to be cut down to 2.4 m wide, although special sway damping might be enough.

    Another solution would be redesigned bogie-frames, for $50million or so. (Information from Siemens: I wrote to them some years ago)

     
  25. John Rankin, 22. August 2019, 15:53

    @GlenSmith: thank you for making the time and effort to answer my questions. I appreciate this. A couple of quick points:

    – Regarding light rail platform design. The preferred design for on-street systems is light rail runs down the middle of a wide street, with a central platform for people getting on and off. So not the artist’s impressions shown in the LGWM documents (which are just bus stops); as you say, real stations, but for low floor vehicles.

    – The next stage of LGWM could fix things, if they actually go to the market with a technology-neutral request for expressions of interest. Public consultation absent information from the marketplace of suppliers is of little value in my view. The problem may be that NZTA is used to going to the market for a specified chunk of physical road, but what LGWM is buying is a system that has to deliver a suite of services. Normally, when you buy a system you state the problem and ask potential suppliers how they propose to solve it. You don’t tell them the answer you want. I fear that instead, LGWM is following a high risk path where they decide the answer before finding out what the market has to offer.

     
  26. Ross Clark, 22. August 2019, 20:06

    Oh, what an interesting thread we now have!

    @DaveB. The comment that public transport struggles to connect the ‘through’ or ‘beyond’ markets, is based on:
    (a) observation with the local buses, which for the most part are routed through the city centre. The buses pick up passengers as they come into the city, but most of these get off the bus, even as other passengers in the city get on.
    (b) Neil Douglas’s work on the number of rail passengers who connect to a bus and travel beyond the Wellington CBD. It’s not many, but even with my preferred money-no-object option – an extension of the current passenger rail network into the CBD, via a subway – it still wouldn’t be many.
    (c) An exercise I did, looking at passenger rail flows in a particular suburban rail network. 25 percent of the journeys were within the catchment itself, and 70 percent from the catchment into the central city and return (so, a little more focussed on the downtown than Auckland’s). Only 5 percent were going beyond the central city.
    (d) The issue in Wellington would be the combination of a relatively low frequency in the rail network, and high end-to-end travel time as a result. In the days when I lived in Tawa, getting out to the airport would, on public transport, take around an hour, more or less. A direct train, if that were possible, would still take 45 minutes or so, and have the frequency issues. Outside the peaks, which is when most airport traffic is, I could drive it in 25-30 minutes – and no frequency penalty.

    @Glen Smith. The data for Terrace Tunnel traffic you cite, is this for the peak or all-day? The problem is, where the final destinations are. In either case, it is likely to be far too dispersed to make a dedicated link workable.

    @Everyone. In Edinburgh, we have had a light rail line between the city and the airport for over five years, more than enough to see what works, and what doesn’t. The main issue? The lack of grade separation from the city terminus to Haymarket well-nigh guarantees that trams are bogged down in peak traffic. Those of you wanting to run trams down the Golden Mile, please consider that in practice, they might not be any faster than the current bus services. So yes – I agree with Dave B’s response to Glen Smith @1354 yesterday.

     
  27. John Rankin, 23. August 2019, 8:24

    @GlenSmith: one more thing. You say “Which is why customers hate transfers.” I’m sorry, but this is not true, although it is true in Wellington, as we are generally rational travellers.

    Good transit operators recognise that good connections are right at the top of customer satisfaction metrics. See for example the 2018 annual report of customer satisfaction from Edmonton Transit. Earlier years are also available. I chose Edmonton because I used to live there and visit regularly, the service is very reliable, and there is an extensive trunk-and-feeder light rail network. You will see that 67% of survey respondents are satisfied or very satisfied with their connections; quality of connections and frequency of service are first equal importance.

    I suggest that effective transit means making balanced investments in getting a lot of things right. Connections are just one factor, although a very important one.

    @RossClark: have an upvote. If LGWM chooses to run light rail down the Golden Mile it could be even slower than the current buses. As I understand it, light rail in a pedestrian priority area is usually limited to a maximum speed of 20 kph. So yes, whatever rapid transit technology we use, it needs to go on the Quays or underground.

     
  28. Ross Clark, 23. August 2019, 20:52

    @John Rankin. I would say, from my UK experience, that transfer-based systems only work when you have a high frequency across both sectors in the journey. People do not like waiting at stations, and will sometimes prefer longer journey times if a transfer can be avoided. This is why we have (or have had) longer-range bus services from Wainui and the Kapiti Coast into Wellington – if a single bus journey can avoid two transfers (e.g. the Coast through to Taranaki Street) all the better.

     
  29. Henry Filth, 24. August 2019, 1:10

    When there is all this expertise in public transport lying around ready for harvesting, why have the WCC and the GWRC spent so much on consultants in recent years? Can they not read Wellington.Scoop? Can they not read?

     
  30. John Rankin, 24. August 2019, 8:10

    @RossClark: agreed. As I understand it, this is what LGWM is proposing, at least in principle. The difficulty with too many longer range services is how best to avoid having services that compete with one another for the same riders. I think that is sometimes called The Triangle of Frustration. You can’t please everyone.

    Coming back to the alternatives Glen and others are putting forward, let us hope that a bold supplier will put in an unsolicited bid, as is happening in Auckland.

     
  31. Chris Calvi-Freeman, 24. August 2019, 19:34

    Henry Filth – in fact Roger Blakeley and I brought together a group of very knowledgeable LRT advocates for a series of workshops which eventually positively influenced LGWM’s decisions on the proposed mass transit route.

     
  32. Glen Smith, 25. August 2019, 8:06

    Dave B. I essentially agree with all you say…as the ideal…in theory. But then you have to mix in a large cup of pragmatism based on the real world. What you describe requires going underground or overhead and the financial, logistical and political realities is neither of those are going to happen (and I think can be shown to be unnecessary). If advances aren’t made under the current coalition then the opportunity will be lost because National will get back in and spend all the money on even more roads. The fall back position is to get as close to what you describe as realities allow. You say you want ‘heavy’ rail which is ‘something compatible with what we already have in terms of passenger-train capacity’. The key question is how much across town capacity do we require?, can across town surface supply this?, and if so what does it need to look like?

    Lets assume that at some point the ‘pinchpoint’ has been solved and both the major Hutt and Kapiti lines are running to full capacity at peak times (let’s put Johnsonville aside at present). If you wanted to run ALL trains as though units you would need 4 lines and 8 carriage across-town stations. That not possible in a compact city, but also, in my view, not necessary. All our Matangi fleet and network capacity is gainfully employed servicing the current demand from the northern CBD. If you run all units through from the north then you will be unnecessarily running underutilised semi-full units (requiring extra network design capacity) once these northern CBD commuters exit. So if across town network capacity is at a premium, then across town units should only be a subset of total units. ie we should have units to the station AND through units. Since the Matangis are gainfully employed (and I suspect could also supply the Palmerston North run if the sensible decision to electrify was made) and we are expanding capacity we will need new units and these can be made however we like (although perhaps at a slightly higher price than ‘off the rack’ units).

    The potential through ‘market’ (based on Terrace Tunnel traffic and adding likely ‘bleed’ via the Quays/ terrace off ramp etc) would likely be 45,000-50,000 bidirectional (see my comment of May 24 after my ‘Stealth’ article of May 23rd). This is say 25,000 each way and assuming a 40% PT share (perhaps optimistic given Ross’s point that non CBD destinations are more dispersed but I think achievable with a well designed network) gives 10,000/day each way. Averaging this over say 10 hours (so an overestimate but on the other hand not allowing for peaks – although through traffic is much less peaked than commuter traffic – never dropping below 71% of ‘peak hour’ at any time between 7am and 8pm- graph 19 page 17 same data source) is 1,000 per hour each direction. This doesn’t include current station users who choose to get off ‘midtown’ or commuters on any Wellington bus lines we transfer to rail. So let’s double it to 2,000 per hour per direction to be safe. This volume is easily achievable without 8 unit carriages or running every 2 minutes. If we have futureproofed say 90 m long platforms (tight but achievable) that is enough for the equivalent of a 4 car Matangi units with a maximum capacity of around 750. You would only need one full such train every 20 minutes in each direction to supply likely demand or one fully seated train every 8 minutes. We should of course be planning well into the future and I think we can achieve a much better rail corridor than that.

    You say “we need a dedicated rapid-transit corridor free from intrusion by any other traffic” and that this “rules out any form of on-street running or level-crossings”. If a rail corridor is dedicated, what difference does it make if it is running down the middle of a ‘street’ (such as the Quays) or separately? And, as Greenwelly points out, our network has multiple at grade level crossings where cars and rail cross – the key is to manage these well. I think a very high quality dedicated (but not fully segregated) surface corridor that would supply the high capacity you seek and which could support units suitable to ‘trackshare’ on our current network is very achievable given the right route (not the Golden Mile or Newtown) and design, and that this could be more than adequate to supply seamless across town demands for the foreseeable future.

     
  33. Glen Smith, 25. August 2019, 8:26

    John. I had a look at the Edmonton system and I not sure it stands up to the ‘trunk and feeder’ design you say. The proposed LRT network has a classic ‘connective radial’ design. Nowhere could I see 4 lines ending just outside the CBD and continuing as one line of different mode across the CBD.
    The buses are a bit more complicated – almost 400 lines . However looking at lines on their line finder shows most are ‘direct’ services either just to the CBD or as ‘radial connective’ lines across the CBD. The trunk and feeder lines seemed to be all in the outer areas (the Edmonton ring road diameter is around 20km).
    So Edmonton seems to have a logical design. ‘Radial connective’ for major lines (LRT), ‘direct’ buses close to the CBD (within about 7km or so) and ‘feeder’ buses on the periphery. Don’t you think we should do the same?

     
  34. Glen Smith, 25. August 2019, 8:33

    Guy. I was meaning rail from rail (which is what I assumed Mike meant) when I said none of our trains are grade separated – the only place this happens is where the Hutt and Kapiti lines separate. Grade separation of road from rail (the ideal) for a station to airport line would be possible in some places (either side of the stacked multipurpose 2nd Mt Victoria tunnel, at the Ruahine St/Wellington Road intersection, and by the airport). Across town wouldn’t be possible but could be minimised with the right route and rationalisation of Quays side road turns.
    Grade segregation of pedestrian from rail is also ideal. This is something LGWM doesn’t appear to have attempted and based on their artist impression they seem to have removed the City to Sea bridge with people crossing the Quays in large numbers, impairing the efficiency of all modes.

     
  35. Glen Smith, 25. August 2019, 8:49

    CCF. The LRT may have been knowledgeable (if unidentified) advocates but experts will usually do what you want/ask them to do. If you say “tell me what you think is the best single ‘trunk and feeder’ LRT from the station to the airport via Newtown” then that is what they will do. If you say “can you come up with a range of costed options on the best designs for a transport network across the whole region” then that is what they will do.
    The question they were asked is very clear from the minutes of the closed meeting on 5 June 2018. The task was identified as “the preferred route for mass transit between Wellington Railway Station and Wellington International Airport” and “the first stage of mass transit from the railway station to southern Newtown/Zoo was assumed as a given”.
    Who decided that these experts should only look at a single line between the Station and the Airport?. And who decided it should go via Newtown? Not the public. Was it you who made these decisions? And if so on whose authority and on what logical basis?

     
  36. John Rankin, 25. August 2019, 11:40

    @Glen: in Edmonton, suburban light rail stations are hubs for feeder bus lines, just as is proposed for Wellington. The first light rail line terminated in the city centre. It took about 25 years to create the first cross-city radial line. In large part this delay arose from the costly decision to run underground through the city centre. I expect in the fullness of time Wellington will extend the rapid transit line north of the railway station to Kaiwharawhara and the northern development areas, creating the first radial rapid transit line.

    The north west and north east lines in Edmonton converge in the city centre and track-share on the southern part of the line. This has been a woeful experience. The signalling system has proved so unreliable that speeds and frequencies are still not what they are supposed to be. In short: the 21st century signalling on the new line does not talk properly to the 20th century signalling on the old line. The signalling of course talks to the public information displays, so these haven’t worked properly either. I understand the contractor has cut its losses and walked away. Good luck getting 4 lines to track share through the centre of Wellington. I hope you have budgeted to upgrade the signalling systems on the existing rail lines.

    The older north-south lines use high floor vehicles; the new east-west lines being built use low floor vehicles. People can transfer between lines in the city centre. Sometimes leaving old technology behind is the right thing to do, so you can move forward.

    Final point: buses and trains run on time, showing up when the timetable says they will, come rain, snow, or sunshine. If we are going to offer people connections, services have to actually connect (and be frequent, as @RossClark reminds us).

    I support your comments about grade separation of pedestrians and cars from rail. Vancouver and Calgary do this well, but I doubt that their approach would fly here. You mention the Quays; see also what LGWM proposes for rapid transit at the Basin Reserve. Surely we can do better than this?