top of page


Putting heads together will cure this headache.

The design of any intersection should take into account the type of traffic using it,
the routes that they take, the limitations of the site,
and the budget to achieve the optimal result.

The NZTA's website dedicated to improving Warkworth's intersections features the Hill Street intersection on its header but focuses no works that will improve the intersection.

Please click on the image to visit the website.

 The NZTA have set aside a budget of between $20 million and $100 million solely to improve traffic flow at Warkworth intersections. While some work has been completed for the section of State Highway 1 between The Grange and Whitaker Road, little has been done to address the issues at Hill Street.


 While a Western Collector road is necessary to facilitate future urban growth, by even NZTA's estimates the road will quadruple traffic along Hill Street. It will not act as a bypass of Warkworth but will allow some connection between future industrial areas and the end of the tollway.

 Effectively, NZTA funding the Western Collector will subsidise developers of the Southern Growth Cell. The Matakana Link Road will subsidise developers of the Northern Growth Cell.

 Don't just take our word for it, work it out yourself. The Auckland Council has a Development and financial contributions Estimator. The Auckland Council will collect $6,514 per additional lot in transport financial contributions. For the Northern Growth Cell's 1900-2200 proposed lots, it will collect between $12.3 million and $14.3 million. For the Southern Growth Cell's 3700-4300 proposed lots, it will collect between $24.1 million and $28 million.

 Currently, the proposed Northern Motorway extension will be a tollway, which means that there needs to be a free alternative state highway. In the future, if the tollway or the legislation is changed, State Highway 1 through Warkworth could be downgraded to a local route, meaning that the administration of the Hill Street intersection could be tranferred to Auckland Council. In other words, no longer the taxpayers' problem but rather ratepayers'. Auckland Council should therefore make sure that every cent of the NZTA's budget achieves its purpose.

 The future growth of infrastructure is an Auckland Council problem. The council needs to decide what it wants and build it NOW. 

 This page discusses the design principles that could lead to a possible long term solution. Surprisingly, the most expensive option has a potential cost that is only a fraction of NZTA's budget.

Driver Behaviour
A straight line is the shortest distance between
point A and point B.
Involve a turn and the dynamics can get more complicated.

Here is a diagram where the thickness of the lines denote the proportion of State Highway 1 northbound traffic that follows certain routes at the intersection during an average day.

Here is a diagram identifying the eight destination points at the Hill Street intersection with 7 options for each. With the exception of right-turning traffic into Hill Street, there are 55 point-to-point combinations. 

 The most important element of any intersection design is taking into account those who use it. It must be simple and it must be intuitive.

 When there are multiple connecting roads at an intersection, complication of give ways and signalling can cause delays. Drivers need logical and predictable layouts.

 As the Hill Street intersection is effectively five intersections in close proximity, to improve the traffic flow the connecting roads can be realigned so that intersections are merged, moved further apart, or avoid each other.


 One alternative method to merge intersections is a roundabout. Roundabouts can also improve the assessment stage as drivers only give way to other drivers already on the roundabout. Once on the roundabout, drivers have right of way until they depart the intersection.

 Key elements of the design of roundabouts are the size of the centre island and the distance between approaching and departing lanes. The larger the distance between a departure lane and approach lane, the more time a driver has to make an assessment.

 Intersection design must taken into account the traffic flow of connecting roads. A steady stream of traffic can cause long queues for connecting roads with a much lower flow.


 To the right is a diagram of State Highway 1 northbound traffic. The majority of traffic turns right. The problem with any right of way in favour of that flow of traffic could be at the expense of any through traffic in the opposing direction.

 Merging traffic can complicate flow. Traffic from Matakana merges with Sandspit, Millstream Place, Elizabeth Street, and Kowhai Park traffic before the State Highway 1 intersection.

 The traffic flows at the intersections are different at various times of weekdays, weekends, and holidays. Northbound and southbound traffic, types of traffic, and turning routes for each road vary significantly throughout these periouds. 

 The overall intersection(s) needs to be examined. We have calculated the traffic routes and peak flows for weekday, weekend, and holiday morning and afternoon peak flows and tested those flows on multiple intersection designs.  

Construction Methods
Construction disruption is no excuse to do nothing.
The design of the intersection must also include a seamless implementation plan.

This GIS map shows the road reserve width,  public spaces, and utilities.

Please click on image to have a closer look.

 Any project management must include construction methods and site specific traffic management plans (SSTMP) to minimize disruption to already congested traffic. The preferred  design should include as much construction off highway as possible to minimise disruption. 

 The road reserve and neighbouring publicly-owned reserves provide space to improve the intersection. It can be a more cost-effective option to use the existing Crown-owned land than use the Public Works Act to acquire private land to build a new road elsewhere.

 Prioritisation of works is also an issue. Whilst the intersection with Hill Street receives the majority of the focus of this site, we recognize that any reconfiguration must be part of an overall and synergized solution.

 The stormwater under Sandspit Road near the intersection has reached capacity twice in 2016 - the water flooding Kowhai Reserve carpark and flowing down Elizabeth Street. The Matakana Link Road, which is in the same catchment area as the Hill Street intersection, will seal over 2 hectares of land that previously absorbed rainfall. Furthermore, roads, driveways, and roofs of an additional 2200 dwellings in the Northern Growth Cell will increase stormwater flowing through the Hill Street intersection.

 Considering the amount of Sandspit Road traffic and the lack of alternative routes, to upgrade the culvert a lane-by-lane 'cut and cover' method similar to how cattle underpasses are laid under arterial roads throughout the country could be a possible solution.

 The culvert replacement could provide an opportunity to improve pedestrian access between Kowhai Park and Elizabeth Street. Precast sections, similar to those illustrated, could provide pedestrian access and also overflow for the stream in extreme weather events. 

 Such precast methods can also be used to improve traffic flow at the Matakana/Sandspit Road intersection.

 Sandspit Road descends from the intersection with Matakana Road whereas Matakana Road ascends a spur from the intersection. Within 50 metres of the intersection, there is an 8 metre elevation difference.

 To avoid the southbound Matakana Road/northbound Sandspit Road congestion, an underpass could be laid under Matakana Road allowing northbound Sandspit Road traffic an unobstructed flow. The four sections can be laid individually allowing both lanes of Matakana Road to not be interrupted.

 So far, the stretch of State Highway 1 between The Grange and Whitaker Road has been the focus of works. This has resulted in widening of the seal and a partial dual carriageway than narrows before the Mahurangi River bridge.

 Replacing the Mahurangi River bridge has been considered by some as a contributing factor to the Hill Street intersection congestion problem. While currently we don't assess it to be a contributing factor, in the future it could be with the addition of 4300 dwellings in the Southern Growth Cell wanting to use the tollway.   

 Replacing the Mahurangi River bridge with a four lane solution is possible without disrupting traffic during construction. To the right is an animation showing one such method that it can be done, similar to how the Newmarket Viaduct was replaced.


Case Study:

SH50/Meeanee Road Interchange, Napier

 The Meeanee interchange is one of 19 built in the past decade, all costing more than $5 million. More than half of those interchanges have traffic volumes lower than State Highway 1 traffic through Warkworth. 


 The State Highway 50 intersection with Meeanee Road in Napier was a level intersection controlled by traffic lights. Meeanee Road is an arterial between the Central Business District of Taradale (a satellite suburb) and Meeanee (horticultural hinterland).

 SH50 between Hastings and Napier is also known as the Hawke's Bay Expressway. The expressway starts south of Hastings and terminates close to Napier's airport. Over time, interchanges and large roundabouts have been built along it. Traffic on the single carriageway distributor, however, is only as fast as the slowest vehicle on it. Tractors travelling <30km/h frequently causes congestion.

 The problem with the SH50/Meeanee Road intersection was that there were several fatal accidents due to motorists running red lights at speed and colliding side-on with other vehicles. Although these fatalities were not attributed to the intersection's design, NZTA factored these accidents into their prioritisation criteria.

 State Highway traffic volume at the SH50 (22,000 Average Daily Trips) intersection is similar to SH1 traffic at Warkworth (22,206) and the local traffic volume of Meeanee Road (14,652) is lower than the four road connections east of Hill Street (15,500). Compared to Hill Street, congestion was not a major problem either. NZTA, however, allocated a budget of $5.35 million in 2005 to build a diamond-shaped interchange, complete with Art Deco theme to replace the traffic lights.


 In this section we examine various types of intersection format, look at where they have been implemented, and applied the format to the intersection with Hill Street. We accept that we don't have all the answers (or the resources) but our aim is to promote discussion through sharing information and ideas.

 In our animations, we have used NZTA's traffic data and the same modelling software that NZTA and their consultant engineers use (Paramics). We have also tested the results against other software, such as AIMSUM and VISSUM

Traffic Light Formats

The above reconfiguration of the intersection improves the operation of a SCATS traffic signal intelligent transport system.

Please click of the image to magnify.

 Above is the "K" intersection design. By allowing for longer turning lanes, sensors can be further from the intersection and provide information so phases can be more dynamic. By turning all eight points into one large intersection, the risk of short queues blocking other intersections is eliminated.

 The design puts five roads into the State Highway 1 intersection instead of four. The distance across the intersection also doubles, taking longer to clear.

 The downside of such a design is that there are more phases in each cycle to accommodate as many of the 49 turning combinations. Traffic could be waiting longer proportionate to the amount of time that there is a green light.

 Clearly, this involves more queues and more lanes. Queues are already the problem.

This solution is built on the existing roadway. This is obviously a problem.

 Overall, comparing the costs of a proposal with the benefits, this is the least desirable option.

 There have been various tweaking of the traffic light sequences over the past decade. The intersection applies a program called SCATS, which is an acronym for Sydney Coordinated Adaptive Traffic System. Also used on the Auckland ramp meter system, here is a description of how it works:


SCATS primarily manages the dynamic (on-line, real-time) timing of signal phases at traffic signals, meaning that it tries to find the best phasing (i.e. cycle times, phase splits and offsets) for the current traffic situation (for individual intersections as well as for the whole network). This is based on the automatic plan selection from a library in response to the data derived from loop detectors or other road traffic sensors.

The system uses sensors at each traffic signal to detect vehicle presence in each lane and pedestrians waiting to cross at the local site.  The vehicle sensors are generally inductive loops installed within the road pavement. The pedestrian sensors are usually push buttons. Various other types of sensors can be used for vehicle presence detection, provided that a similar and consistent output is achieved. Information collected from the vehicle sensors allows SCATS to calculate and adapt the timing of traffic signals in the network.(link 1, link 2)

  Any intelligent transport system is only as effective as the quality of information that it receives. The locations of the sensors are crucial to providing information on turning directions and the sooner that information is available assists with the timing of sequences.


 The problem with the intersection at Hill Street is that there are short turning lanes and sensors are close to the intersection. Considering the many different peak flows and fluctuations in those flows (such as for events), SCATS phases can easily be out of synchronisation with realtime flows.

 To improve the effectiveness of the SCATS system, low cost suggestions such as restricting turning into and out of Hill and Elizabeth Streets have been proposed. A more expensive suggestion involved a redesign of the intersection, involving longer turning lanes, traffic signals for the Matakana/Sandspit Road intersection, and incorporating Elizabeth Street into a "K"-shaped intersection with State Highway 1 and Sandspit Road.

Large Roundabout Formats

 Roundabouts offer solutions that have simple give way to right decisions. Large roundabouts can take on various shapes, sizes, and configurations. The benefit of a large size is that it allows traffic to negotiate it more quickly whilst the larger distances between approaching and departing connections allow for the smoother flow of traffic due to longer response times. In other words, traffic approaching the roundabout can see whether traffic on the roundabout will depart at nearest connection.

 Two lane roundabouts must be large to flow smoothly.

 Some large roundabouts can operate more like a one-way system. In Wellington, the Basin Reserve cricket ground is a 600 metre multi-lane perimeter roundabout with four major arterial roads and many minor connections.

 Large oval-shaped roundabouts can merge two clusters of connections and provide a more logical layout.

 Take, for example, the State Highway 20/Queenstown Road peanut shaped roundabout in Onehunga. There are two clusters of connecting roads with ramps to and from a motorway. The predominant flow is the left turn from Queenstown Road onto SH20 towards Mangere Bridge and the airport. Intermittent flow turns right from the SH20 offramp north onto Queenstown Road.

 The traffic patterns at Onehunga are not too dissimilar to those at Hill Street. Sandspit Road predominantly turns left onto a State Highway 1 and a significant but intermittent northbound State Highway 1 traffic turns right onto Sandspit Road. The major difference, however, is that right turning traffic in Onehunga overpasses the state highway.

  A large roundabout format can be applied to the topography of the site around the intersection with Hill Street in a variety of ways. The Kowhai Park reserve's carpark provides space for a large roundabout but is limited by the neighbouring deep-banked stream. The Sandspit Road corridor across the stream, the Matakana Road spur and the Elizabeth Street and Millstream Place terraces do not provide a conventionally flat area for a roundabout.


  To maximise the efficiency of a roundabout, the connections need to be spread out over a larger area. One way to achieve this is to separate each direction of traffic along Sandspit Road and have approach and departure connections some distance apart. This would require a new crossing of the stream some distance to the north.

 Two methods of large roundabout formats improve the flow of the predominant state highway traffic whilst providing for the intermittent bursts of flow of arterial and local roads are proposed below.

 The concept to the right is nicknamed "The Pill." The elongated roundabout provides a good distance between state highway approach and departing lanes whilst maintaining a very manouvable format. Distances between Elizabeth Street and Sandspit Road approaches are also maximized.

 An offramp serves Matakana Road, Sandspit Road, and Millstream Place. The distance between this offramp and the Sandspit Road onramp improves the flow onto the roundabout.

 The Pill has one pinchpoint, however. The right-turning traffic at the northern end of the roundabout can hold up southbound state highway traffic. If this becomes a problem, it can be controlled by signalised enhancements at intersections with Whitaker Road and Hudson Road.


 The major benefit of a large roundabout format is that it can be built with minimal disruption as it effectively is built to the side of the existing intersection.

 Overall, elements of a large roundabout concept has its merits.

Double Roundabout Formats

 Another intersection format relevant to clusters of connections with a variety of traffic flows is the double roundabout format. The NZTA has effectively applied the format throughout the country, including Napier and Mount Maunganui as shown in the examples below.

 The Mount Maunganui SH2/SH29A intersection at Baypark is a main state highway with a branch from one main roundabout to another smaller roundabout serving the harbour bridge arterial and a local road serving the stadium. The main roundabout has two lanes, which improve flow for through and right turning traffic. 

 Napier's SH2/SH50 intersection is effectively a main state highway with two roundabouts in series providing branches for arterials. Being in a 50km/h zone, the roundabout sizes are smaller and longer vehicles use both of the two lanes to manoevre right turns.


 For the Baypark intersection, the major pinchpoint is the SH29A right-turning traffic onto SH2. Although the traffic on SH2 and SH29A have a lower traffic flow than the intersection with Hill Street, NZTA forecast considerable growth and have invested $100 million to replace this roundabout with an interchange and motorway upgrade.

 Applying a double roundabout format to the intersection with Hill Street would look like the model to the right. A two lane roundabout and a single lane roundabout would work according to the animation.

 The benefit of the dual roundabout design shown is the smaller footprint and cost. The kerbing is very similar to the existing intersection, all works are within the existing road reserve, and construction can be completed within a short timeframe. The mountable centre islands have a wider turning radius than the existing intersections.

 The main limitation of the design is the distance between approach and departure lanes. Southbound Matakana Road traffic effectively has to stop to assess whether northbound traffic is turning right into Sandspit Road or heading straight ahead. Similarly, southbound state highway traffic could be held up by local traffic turning right into Sandspit Road.

 While double roundabouts allow for clusters of roads at different levels, each roundabout needs to be built on flat sites otherwise vehicles can't safely negotiate the corners. Earthworks would be required at the Matakana/Sandspit Road roundabout to achieve this.

 The design works well up to 80% of weekday peak flow. Beyond that level, one possible solution could be to install SCATS queue sensors that trigger traffic lights to let queued traffic through.

 As a short term solution, this solution works 45% better than the current format. Against 2026 and 2031 forecasted traffic flows, in the long term, just like the Baypark intersection, a double roundabout won't be able to cope with traffic growth.

Interchange Formats

We've adapted The Pill with overpasses and underpasses, which has doubled capacity.

 On a cost and traffic capacity scale, an interchange is the ultimate solution. Instead of traffic giving way to each other, traffic is uninterrupted by either underpassing, overpassing, having their own parallel lane or merging lanes.

 Interchanges are effective when there are two distinct patterns of traffic between two groups of connecting roads. They are also effective in tight areas that serve many connections.


 The Dowse Interchange between SH2 and Hutt Road in Petone is an example of a state highway with a minor local road on one side and a major arterial and local roads on the other side. The interchange replaced traffic lights where queues for right-turning traffic often extended past the previous two intersections.


 The SH2 traffic at Dowse (32,000 ADT) is what NZTA forecast the State Highway 1 traffic at the Hill Street Intersection will be in 2031. Current Hutt Road traffic, however, is the same as what Sandspit Road traffic will be in 2022.

 The Dowse Interchange is a good example of grade separation. Motorway, arterial, local, cycling, and pedestrian traffic all have separate zones.

 The natural topography of the Hill Street intersection environment lends itself to an interchange. With the exception of Elizabeth Street, all roads descend towards the intersection. Taking advantage of the crest and orientation of the northern and southern approaches of the state highway can form a fluid curve and elevation for an overpass. Taking advantage of the difference in elevation between Matakana and Sandspits Roads lends itself to an underpass.


 Incorporating The Pill's design features into an interchange layout separates local from highway traffic, solving the conflict of right turning traffic towards Sandspit/Matakana with southbound highway traffic.

 An interchange is a solution to a dominant flow. When the tollway is constructed, Hill Street intersection will become a 6-way distributor and the dominance of State Highway 1 will be reduced in the short term. In the longer term, however, as the Southern Growth Cell is developed, traffic along State Highway 1 will grow to higher volumes by the time that the tollway is completed.

 The benefit of The Pill model is that it can be the first stage of an interchange design. It would be prudent to plan ahead by building the interchange sooner rather than later. 

 Overall, holiday traffic is catered for by the Matakana Link and Tollway. The problem at Hill Street intersection is distribution during normal operation. Holiday traffic just compounds the distribution problem.

The Pill concept is a model that our team is continually finetuning to ensure the optimal flow whilst minimising cost and disruption. While the emphasis of the design is the large roundabout, an  interchange overpass can be added at any stage without disruption.

Fixing Hill Street last will produce disruption.
Fixing Hill Street first will produce benefits for other works.

 There are currently four sets of traffic lights along Warkworth’s urban stretch of State Highway 1. The Tollway, Matakana Link, and Western Collector will add three major intersections. The Matakana Link will also add a major intersection to Matakana Road and potentially Sandspit Road in the future.


 Of the seven current and proposed intersections between the tollway and McKinney Road, four are T-shaped, two are X-shaped, and Hill Street has 8 connections.

 Currently, State Highway 1 traffic to Hill Street is controlled by the Hudson Road and Whitaker Road traffic lights. These gaps in flow will benefit a roundabout format at Hill Street for local traffic. If Hill Street is upgraded to an interchange, however, there will be uninterrupted flow through the intersection.

 The order that projects are built is important. Get the order wrong and there is disruption. Get the order right and there could be flow-on benefits.

This diagram shows the increase of Warkworth's State Highway 1 intersections from four to seven.

Please click on the side images to enlarge. Hudson, Whitaker, and Woodcocks Roads intersections include suggested improvements. 

 There is potential to reconfigure other t-shaped intersections to through traffic to have unhindered lanes. Southbound State Highway 1 traffic at Woodcocks Road, Northbound State Highway 1 traffic at Whitaker Road, Southbound State Highway 1 traffic at the Tollway junction, and Northbound State Highway 1 traffic at the Matakana Link junction will not need to stop for traffic lights or give way to other traffic at the intersections. Hudson Road intersection could be reconfigured to a t-shaped intersection so that showgrounds and industrial traffic uses the Matakana Link, allowing southbound State Highway 1 traffic unhindered flow.  

 Effectively, northbound traffic will only need to stop or giveway at four of the seven intersections through Warkworth and southbound traffic will only need to stop or give way twice.


 Fixing Hill Street must be part of an overall solution. The NZ Transport Agency recognizes this when they allocated a budget of $20 million - $100 million to improve all Warkworth intersections to improve safety and journey times from one end of the urban boundary to the other. Fixing Hill Street first will produce benefits for other proposed works.

This page's aim is to show what can be done. We aren't providing the answer.

The answer will come from the NZ Transport Agency and Auckland Transport working together.

Obfuscation can not be an excuse to do nothing. There needs to be focus, ownership, and leadership to make fixing Hill Street a priority. 

- Interchange
Driver Behaviour
- Traffic Lights
- Large Roundabouts
- Double Roundabouts
Case Study
bottom of page