Wellington Scoop

Avoiding congestion by rethinking Melling Bridge plans


by Glen Smith
The New Zealand Transport Agency has put forward three plans to improve traffic flow on State Highway Two at the congested Lower Hutt turn off across Melling Bridge. Improvements at this intersection deserve support since the aim should be to improve the efficiency of all transport modes and to reduce the potential for accident related deaths and injuries.

However the evidence from here and overseas is very clear: road improvements that aren’t balanced by corresponding improvements in public transport lead to increased total car trips and, inevitably, to accelerating congestion.

Professional modelling of the effects of road projects proposed by the NZTA under the previous National Government confirm this effect.

The Opus TN24 Baseline Forecasting report, which forecasts the effects of these proposed road schemes, predicted 50,000 extra road trips by 2041 producing an overall increase in congestion across the city of nearly 90% and a staggering 405% in the am and 434% in the pm to and from the Hutt (see table 6-4 from the report below). It is hard to imagine congestion at this level but that is the modelled prediction.

The closest we’ve come to this sort of congestion was during the 2013 Hutt Rail washout which saw commuters from the Hutt delayed by up to several hours daily at an economic cost of around $1.3 million per day.


To counter escalating congestion, significant improvements in public transport are required. An efficient and attractive service with accurate and predictable timetabling will be increasingly impossible for non segregated public transport, particularly buses on public roads, as congestion climbs. Dedicated corridors are required.

Lower Hutt has a dedicated public transport corridor in the form of the Melling rail line which has the theoretical capacity up to 8 lanes of traffic. However the line is hamstrung by a number of limiting factors which means it is grossly underutilised, with only one train per hour during non peak periods. To minimise future congestion, the aim should be to maximise use of this line.

Probably the most important limiting factor is that the Melling Station is over a kilometre from the main Hutt city centre on the opposite side of the river. Replacement of the Melling Bridge offers an ideal opportunity to rectify this deficiency.

The NZTA plans to move the Melling Station and construct a new pedestrian and cycle bridge that will bring the station closer to the city centre. However major destinations will still be over half a kilometre from the station and on the opposite side of the river, a major deterrent to utilisation. The pedestrian/ cycle bridge is unlikely to be engineered to carry the weight of rail, however the new Melling Road Bridge will be engineered for heavy traffic. The cost of adding a single 3m wide corridor for rail is likely to be insignificant and would allow rail to reach to the very centre of Lower Hutt.

The question then would be how best to utilise this rail corridor.

The Green party proposes, at an unspecified time in the future, to utilise this line for light rail running from Island Bay to Epuni Station. This proposal would appear to face some significant challenges.


A dedicated light rail to Island Bay would be hugely expensive and destructive. A non dedicated light rail line is unlikely to be sufficiently superior to buses to justify the large extra cost, given current transport volumes. In the foreseeable future, Island Bay is logically best serviced by buses.

Rail through Lower Hutt would have to travel through busy multipurpose spaces and would have to be at the ‘light’ end of the rail spectrum. There are major practical issues around running such truly ‘light’ rail units on the same lines as ‘heavy’ rail between Petone and the Wellington Railway Station.

More importantly, there is a limitation on the number of rail units per hour that can be safely accommodated just north of the Wellington station. Some relatively simple alterations in the layout and functioning of the rail corridors in the area between the Railway Station and the separation point of the Wairarapa Line and North Island Main Trunk line could significantly increase unit capacity in this area. However this doesn’t seem to even be on planners’ radars, despite the inevitability of such changes and the logic of undertaking them now. Even if undertaken, any increase in the number of rail units per hour would be more logically used to remove mode transfer at the station for commuters to the Airport from the Hutt and Kapiti lines.

The alternative is to aggregate demand on the Melling/Lower Hutt line onto larger multi-unit trains servicing the Hutt/ Wairarapa Line. Unfortunately the Melling Line joins the Hutt/Wairarapa line mid way between stations near Mill Road in northern Petone, meaning light rail units would still have to travel amongst ‘heavy’ rail units as far as Petone Station. However Petone station and the next station north, Ava Station, are suboptimally spaced at around 2km apart, leaving an area of north Petone poorly serviced.

The logical solution is a new station servicing northern Petone and Maungaraki while acting as a transfer station to the new Lower Hutt light rail.


On the Lower Hutt side of the new Melling Bridge, a range of routes would be possible. The key question here is what purpose rail would be serving. Buses already provide extensive services and the primary aim shouldn’t be to try to replace these since rail can never achieve the coverage of buses, and attempting to do so generally imposes additional transfer penalties on commuters. Instead the aim should be to provide rapid high quality rail based transportation to the Lower Hutt CBD and Queensgate Mall from Wellington and Upper Hutt on a fully dedicated corridor, without any ‘steel to rubber’ transfer disincentive.

The key destination in Lower Hutt should be the main Queensgate bus stop in Bunny Street. The aim would be to reach this with minimal interaction with main road routes and minimising loss of parking. The diagram below shows my suggested route.


Rail commuters from the north wouldn’t want to travel around to Petone to access Lower Hutt. The sensible solution is to continue the light rail corridor to a station to the east. The largest closest station which is serviced by all trains including express units is Waterloo Station. This is a straight run along Knights Road from Lower Hutt and Queensgate. The best overall route therefore in my view would be as shown below.


This route is only 4.8 kilometers of which only 2.3 kilometers is new line, most of this being a straight run down Knights Road likely using the central median strip for width.

Congestion isn’t solved by building new roads. It is solved by providing high quality public transport that encourages commuters out of their motor vehicles for trips where cars aren’t essential.

Despite this fact, history has demonstrated that our ‘Transport’ Agency has an illogical bias towards road based solutions, a prejudice that only serves to increase congestion in the longer term. They appear willing to proactively plan high quality public transport only when severe congestion, such as seen in Auckland and increasingly here in Wellington, forces them to do so.

I encourage the NZTA to reconsider their Melling Bridge proposals and to seize the opportunity to establish a truly high quality, rapid, dedicated, rail based public transportation corridor to the Lower Hutt CBD.

By doing so they will help to avoid the rapid and hugely expensive increase in congestion that is projected to increasingly cripple our region in the coming decades.


  1. Andy Mellon, 10. May 2018, 18:57

    I don’t see the benefit of light rail into the centre of Lower Hutt in comparison to extending the heavy railway line to Kelson/Belmont.

    Melling has grown to be very successful as a commuter rail starting point. An extension to Kennedy-Good along the old railway workings would provide the same service to an extended range and have the maximum benefit in taking vehicles off SH2 at peak times. Extending the line into Lower Hutt is much less likely to take vehicles off SH2.

    Furthermore, why aren’t uses for the Gracefield line being looked into? A commuter Park & Ride service from Gracefield to service Eastbourne/Bays/Wainuiomata/Moera commuters would also help to alleviate traffic flows. The rail’s just sitting there, diesel units could be used for a pilot service to see if the demand exists before later electrification. This would also have the benefit of increasing capacity through Southern Hutt.

    Hutt railways could and should be extended, and the designs for the Melling interchange provide an opportunity to provide a rail extension at the same time. I just don’t see the need to provide service to an already well serviced area that won’t have much impact on peak time journeys. Extending to Kennedy-Good provides rail access to an under-serviced community.

  2. C Lawless, 10. May 2018, 21:10

    The last time Lower Hutt Light Rail was brought up it resulted in Jville to the Airport being side-tracked. So beware of confusing planners and ending up with nothing.

  3. Glen Smith, 12. May 2018, 8:01

    Andy Mellon. The purpose of the extension is so that rail services the Lower Hutt CBD, which data shows it fails to do at present and which represents a major deficit in our rail network. The 2017 Rail Survey shows that 392 people boarded at Melling Station during the am period, a useful number. However the Rail Passenger Survey showed that 62% arrived by car and only 27% by foot- this is just over 100 people or around two busloads. Compare this with the commuter traffic to the Lower Hutt CBD. Queensgate Mall alone quotes a footfall of 8 million per year or an average 21,000 per day requiring around 2,000 carparks occupying prime downtown commercial space.
    Melling Station effectively functions as a rail line to a park-and-ride carpark which could almost be anywhere but just happens to be across the river from the Lower Hutt CBD. The aim would be to retain and expand this park-and-ride capacity (by continuing to add car spaces as demand dictates, even if this involves going up) but to improve it to actually service the commercial area. Realistically this means the line has to penetrate the Lower Hutt CBD.
    This proposal involves a transfer to overcome the practical problems with running ‘heavy’ rail into the Lower Hutt downtown area, or with running ‘light’ rail to either Wellington or the Upper Hutt area. This incurs a rail to rail transfer disincentive which Neil Douglas’s research (previously referenced in a Scoop article) estimates at around 9 minute ‘pure’ penalty, a lot less than a rail to rubber penalty of 17.5 minutes. With a dedicated corridor and a Hutt Rail service that runs 99% to time, a sub 3 minute ‘walk and wait’ time should be achievable. The research shows almost all Melling Station users are working commuters (92% in the 22-64 age group, 89% making the return journey at the end of the day) and these would soon get into a transfer routine or, if this is a major impediment, use the Petone park-and-ride facilities
    Looking at the Kennedy-Good option, I assume the station would be near the western end of the Kennedy Good Bridge. People again would likely arrive predominantly by car or walking. People are generally prepared to walk about 500m. The walking catchment therefore is, to the east, only empty river reserve and to the west only the very lower most areas of the Kelson and Belmont hill suburbs. The station would therefore function as almost exclusively a park-and-ride station which could be anywhere (including at Melling a couple of minutes drive down SH2).

    C. Lawless. This proposal is completely separate to any Johnsonville line proposal. If planners can’t multitask and consider both concurrently one has to ask if they should be planners.

  4. Andy Mellon, 12. May 2018, 9:20

    People using Melling will not use the Petone park and ride facility. If you take this away, then you’re exacerbating the issues you’re purporting to solve. Providing Park and Ride facilities still takes a vast amount of commuter vehicles off SH2 during peak traffic.

    How much peak traffic flow is a Hutt CBD station going to remove? Who is commuting from Hutt CBD to Wellington (or vice versa) during the peak time? I don’t see how a Hutt CBD station improves peak time traffic flows in the slightest. People will want to travel to the Hutt CBD, but these journeys if by car are against the main traffic flows.

    Melling’s success as a park and ride is because it’s at the base of Maungaraki. Drive or walk down, you don’t need to go along SH2. I can’t believe that you’re suggesting we add these park and ride commuters to SH2 morning traffic and tell them to trudge along to Petone. Just because you put a station in Hutt CBD, doesn’t mean people will catch the train there. Look how much worse traffic in Petone has got with the big box retailers; all vehicle traffic – hardly a pedestrian to be seen – despite being close to the station.

  5. Roy Kutel, 12. May 2018, 12:16

    Hmmm, I don’t think I’d read too much into the GWRC survey, as there are some very strange statistics in it. For example, half of tertiary students are over 26! Only 10% of rail passengers were under 27 (probably because young passengers weren’t handed a questionnaire or they refused to complete it) whereas surely it must be more like a third. And only 5% access rail by bus (surely it must be double that since if it is only 1 in 20 it’s an indictment of GWRC integrated bus/rail planning). And these are just three examples. Hopefully future planning is not being based on these silly stats. Thanks Glen for alerting me to them.

  6. Glen Smith, 12. May 2018, 20:17

    Andy Mellon. Please reread my comment. I don’t advocate ‘taking away’ Melling park-and -ride but expanding it. However rail should also penetrate the CBD. This rail is aimed to increase PT share and reduce overall congestion across the Hutt by offering the tens of thousands of people who travel to and from the Lower Hutt CBD each day the option of a high quality dedicated extension of the existing extensive rail network – not just to reduce SH2 congestion into Wellington.
    As you say congestion sadly continues to increase not just in Petone but across the region. By your logic we therefore shouldn’t bother trying to offer people high quality PT options. I would be interested in your alternative suggestion for reducing escalating congestion.
    Roy Kutel. The response rate is low but the sample size is still large. The passenger counts were undertaken by observers so will be accurate. There is no particular reason why the response rate of park-and-ride commuters should differ from that of commuters arriving by other modes so no reason to alter the conclusion that Melling functions mainly as a park-and-ride station and fails to service the Lower Hutt CBD.

  7. Alana P, 13. May 2018, 8:36

    Glen, Roy is right. The sample is not large for a profile survey and with no adjustment for differential interception rates (eg young people close to zero) the estimated profiles will be biased. Young people are more likely to walk or catch a bus for example. So beware in using it.

  8. Farmer John, 13. May 2018, 9:32

    Shows how little the GWRC cares about the Wairarapa! Just 13 surveyed in the AM peak, none in the of off-peak and only 14 going home in the PM peak! Its a joke! How can you work out access requirements at Featherston when you only have 4 respondents? The survey is another indictment of the GRWC. They need to be replaced with a Public Transport Authority so the people of the Wairarapa can have the semblance of a decent service planned and operated by professionals.

  9. Glen Smith, 13. May 2018, 9:51

    Alana. I am familiar with analysing research- I have been doing it for 40 years. The survey can’t be faulted on sample size- the 95% confidence interval is 2%. It could be criticised on self selection bias but as stated there is no reason to suspect this is significant.
    Anyway this survey only confirms what is obvious to the casual observer which is that Melling station doesn’t service the Lower Hutt CBD (where are the hundreds of people crossing the Melling Bridge). It is unclear why you would want to spend your time trying to deny this rather than exploring options to solve this deficiency.

  10. Farmer John, 13. May 2018, 10:06

    Just shows how little the GWRC cares about the Wairarapa – just 13 surveyed in the AM peak, none in the of off-peak and only 14 going home in the PM peak! Its a joke! How can you work out access requirements at Featherston when you only have 4 respondents? The survey is another indictment of the GRWC. They need to be replaced with a Public Transport Authority so the people of the Wairarapa can have the semblance of a decent service planned and operated by professionals.

  11. Neil Douglas, 13. May 2018, 10:28

    Glen, Alana, Roy, this paper on sample size and response rate may be of interest. By coincidence the paper refers to a similar Wellington rail survey done in 2004 but done for the train operator not the regional council. The sample was four times larger with every weekday and weekend train counted but corrective measures were still required because of refusals, in order to get a reliable profile.

  12. Andy Mellon, 13. May 2018, 13:57

    Glen – please read my comments. I have offered an alternative. Extending the Melling line to Kennedy-Good will get more peak time vehicles off the road than light rail to Hutt CBD. I’m an advocate for better public transport, I just don’t think the Hutt CBD link is going to reduce peak time congestion.

  13. Glen Smith, 21. May 2018, 7:57

    I see NZTA Wgtn has added a twitter post. It is good to see then actively consulting with the public. The question is if they are listening or whether this is sham ‘consultation’ as we saw with the Basin flyover.
    Congestion in Wellington CBD is predicted to decline to LOS (level of service) E or F over the next couple of decades. Lower Hutt CBD congestion will inexorably follow suit. Avoiding this requires high quality dedicated PT corridors. Adding these is difficult and requires long term strategic planning- the rail corridors that service our region date back to the 1880s. The time to think about this is now when transport changes for access to the Lower Hutt CBD are being planned that involve the dedicated Melling rail line.
    Public transport funding will be stretched in the next few years but plans can be futureproofed. It would be nice to see some modelled and costed options for future Lower Hutt light rail. The bare minimum should be reserving a dedicated rail corridor from the proposed Melling park and ride station across the proposed new Melling Bridge to the east side of the river that can be utilised at a future date.