Life and death of the Stranmillis Gyratory

Cycling infrastructure enables people of all ages and abilities to get about safely and confidently by bicycle. It's what we need more of right across Belfast. But it's only half the story to making Belfast a healthier city through active travel.

The other half requires a very uncomfortable conversation. How do we go about making private motor travel less attractive? Can we actively reduce car journeys as a policy goal?

We excel at carrot-based strategies and plans to support active and public transport but no-one ever really talks stick. Bus lanes are a pet hate, but they're just tinkering at the edges of an otherwise untouched road system. Reducing vehicle journeys in the here and now, not in some forever distant future - is that front and centre of anyone's agenda?

Regardless of active travel infrastructure, if a city's street network is entirely permeable for motor vehicles, every other mode has to play second fiddle - and that usually means suppressed to the margins, as walking and cycling are in Belfast.

Simple assessments of cycling success in The Netherlands sometimes miss this point - yes, the dense cycling infrastructure is of fundamental importance, but how roads are defined and the types of traffic which they cater for also play a crucial role in making certain journeys less attractive by car.

This is why Bikefast is pushing for a Congestion Plan for Belfast to radically reshape how our city plans for mobility - something to shake up the cosy status quo within the Department for Infrastructure and force radical change.

But where exactly to start this uncomfortable conversation? That fell into our lap with a call for action on cycling safety at the Stranmillis Roundabout:

@PSNIBelfastS I have concerns about cyclist safety at Stranmillis R'Bout. Anyone I can chat too ? Ive had 4 close calls in last 3 weeks.

— Niall McFarland (@NiallMcFarland) June 17, 2017

Knowing this junction very well, whacking in a small, calming, mode-separated Dutch-style roundabout was our obvious knee-jerk, junction-focussed fix. Put vulnerable users first and improve safety.

But looking at the wider picture, why is the traffic speed and volume so great here? Focussing on the roundabout is to miss the significance of a weird feature just yards away - a leafy riverside handed over to a large urban gyratory system which attracts a high volume of through-traffic into its orbit from as far away as the Falls and Holywood Roads.

When single junctions have problems, traffic engineers focus on the micro solution and aren't compelled to look at the macro problems. That approach must change to make Belfast a fit, healthy and attractive place to live. If we follow the logic of the problems in this area back to their sources and propose alternatives, it leads to some interesting outcomes.

So welcome to the death of the Stranmillis Gyratory, purging unnecessary through-traffic from an entire city district, and the massive benefits Belfast's riverside and residents could enjoy in its place (long read).

Stranmillis is in a funny spot. It's a desirable postcode, a leafy suburban backwater in Belfast hemmed in by the River Lagan, the countryside of the Lagan Meadows, several university campuses and Botanic Gardens. Stranmillis Road is not a traditional arterial route - it's a 2.3km crescent shaped loop off the Malone Road.

There are two relaxed village-style hubs where cafe culture mixes with art galleries, restaurants and high-end shops:

• Stranmillis Village with its established businesses built on student population
• Lockview with 5A, Cutter's Wharf, the boat clubs and tennis club

The Lyric Theatre provides a cultural hub, nestled below terraced streets filled with students and with a prime view over Botanic Gardens.

And yet, Stranmillis is awash with vehicles busying through, a high proportion with no purpose along the way. Why is this?

The Stranmillis Roundabout feels like it's at the centre of this traffic flow. It's a bloated version of what should be a small residential area junction:

• a double lane roundabout where a single lane would suffice
• five arms creating a dicey injection even for drivers at busy periods
• poor sight lines especially on the countryside arm of Stranmillis Road
• traffic - lots and lots of traffic

Signs point to far off destinations like the M1 motorway 4km to the west. But despite its over-capacity, the roundabout isn't generating that traffic. Look beyond to the river and the culprit is hidden in plain sight.

Two double lane vehicle bridges sitting 300 metres apart - Governor's and King's - and multiple lane roads connecting them on each bank. A clockwise one-way gyratory system spinning traffic into and out of six access points at its corners.

As shown in the diagram below, it's divorced from the main arterial road system, sitting perfectly in the middle spot between Malone and Ormeau. Yet the capacity and permeability of the street system around it leads to several clear and definable - if unsigned and informal - through-routes which have developed over the decades.

Two dominant flows cross over:

1. A route between northeast and southwest bringing significant portions of east Belfast within striking distance of the M1 motorway without having to take the Outer Ring or the M3 cross harbour bridge and Westlink. This axis also provides an alternative route for southwest Belfast traffic accessing the eastern part of the city centre and Titanic Quarter while avoiding arterial congestion.
2. A route between northwest and southeast, a trail from the Falls along Broadway/Donegall Road, up Tates Avenue, along Eglinton Avenue, Chlorine Gardens, Stranmillis Village, Ridgeway Street and across the river to any number of options. This same axis sees traffic from the Saintfield Road direction able to access the University Quarter and city centre while avoiding the Ormeau Road.

It acts as a series of relief corridors for cross-city traffic which (in the immortal words of Department for Infrastructure's Ciarán de Búrca) is "not stopping because they have no business there, going from one side of the city to the other and using it as a short cut".

Moreover, with very few junctions and control points, a lot of the journeys are definitely quicker than the place-appropriate alternative - using arterial roads and ring routes for cross-city travel.

Having identified that a giant turbo roundabout of sorts is dragging vehicle traffic into its vortex, the next step is to find a way to disentangle the flows of through-traffic and local access traffic - eliminating the former while supporting (as much as possible) the latter.

Whatever the future of mobility in Belfast, when vehicle journeys are concerned there should be a simple model for streets and roads which is easy to understand - both for planners and users.

"To the Dutch the most ideal situation is when roads and streets have only one single purpose. To achieve this mono-functionality a hierarchy of roads was introduced.

1. Through Roads for high volumes of fast traffic on longer distances.
2. Local Access Roads from which end destinations can be reached.
3. Distributer Roads which connect through roads and local access roads.

All Dutch streets and roads have been classified (under a legal obligation) and are or will be re-designed to the Sustainable Safety principles by the road managers. This led to areas where people stay (residential areas and areas for shopping/sporting/theatre etc.) and designated space used for the flow of traffic in order to transport people from A to B. Under the Dutch vision these functions cannot be mixed."
Sustainable safety (Bicycle Dutch)

In South Belfast things get a little more mixed than the Dutch would like. However, we can begin to build a similar hierarchy of street purpose using the arterial routes as our distributor roads, the Outer and Inner Rings as our through roads, and everything inbetween as our local access roads.

In the context of the Stranmillis Gyratory problem, we have two distributor roads in the Ormeau and Malone linked to the south by a through road in the Outer Ring. The Stranmillis Road as a looping offshoot of the Malone Road should be serving little more that access traffic - but the gyratory system holds it at distributor level.

How everyone looks when cycling up to Stranmillis Village 😉 "@kiwioconnor What pain looks like #giro pic.twitter.com/QcZv3ELTgi

— NI Greenways (@nigreenways) May 9, 2014

Give road engineers a challenge to devise ways to calm this gyratory system would likely lead to conservative junction fixes - signal phase changes, lane realignment, altered signage etc. Marginal changes which bring short-term effects but would quickly return to the status quo as drivers adapt.

But the nature of the journeys which gravitate towards this area requires more than the tools road engineers can bring to the table. We need to look a full two miles to the north and a mile to south.

A radical approach to pushing through-traffic back onto more appropriate through-roads looks like this..

This is the launch point for southbound traffic from the areas surrounding the massive junction system at Short Strand. From the Ravenhill Road to Ormeau Road the Embankment runs as a full mile of road unhindered by significant junctions or traffic control - quite remarkable in the middle of a decent-sized city.

On the other side of river there is no road between Ormeau and city centre, a legacy of the sprawling former gasworks site and current rail corridor. A hugely important and popular active travel corridor has developed instead, one link in a chain all the way from Lisburn to Newtownabbey.

Ormeau Park itself has been around for almost 150 years, but the embankment road wasn't always there:

"The park was designed by Timothy Hevey and opened to the public in 1871. The opening was marked with a parade from Carlisle Circus through Belfast which attracted a large crowd and finished with speeches in the park. The present day park still roughly follows his design but with several alterations, for example, in the 1920’s, the embankment road cut off the river frontage from the park."
Planning NI

Not only that, there was a beach! pic.twitter.com/K8zTQx4DOl

— Victory Chimp (@VictoryChimp) June 20, 2017

The river slip at Ravenhill Reach, the Ozone complex and riverside apartments all require vehicle access, but there is no other purpose to the Ormeau Embankment other than as a distributor road.

So, we'd lift the road from here to Ormeau Bridge. Gone.

Plans to build a traffic-free bridge linking from the proposed end of the Ormeau Embankment mean there will be a very practical new east-west route opened in place of the north-south vehicle carriageway. Original options for the bridge included an extended bridge over the road to link safely into the park. Cutting the road offers a very simple solution to this problem.

Removing the remainder of the roadway to Ormeau Bridge would reconnect the park with the river bank. A greenway path would nicely replace the road, keeping movements options open for people, if not cars.

The Ormeau Road would benefit from the four-way intersection at the park gates being reduced to three-way. General traffic flows get a little boost while a continuous countrybound bus lane could be protected from the remaining Ormeau Embankment light phase. Cycling and pedestrian movements from Lower Ormeau to the park would encounter no major road crossing.

This immediately gets rid of the short cut to the M1 from beyond Ravenhill, while local traffic still can drive between Ravenhill and Ormeau using either the Park Road route or across the Albert Bridge.

The current quickest way to the Stranmillis Gyratory from the Lower Ormeau is to turn off the Ormeau Road at the northern end of the Ormeau Bridge. It's a difficult junction for anyone trying to turn, with no traffic control save for a toucan pedestrian crossing just beside it.

The residential Holylands streets just off the embankment are well served by access from the Ormeau Road and University Street. Moves over the years to cut the rat run from Botanic through to the embankment have been somewhat successful, but the logical endpoint is removing vehicle access to the river altogether.

The only frontages with access needs sit with the apartment block and cottages beside the Botanic Park playground. Utilising the ramp down from Harrow Street is one option, keeping the residents fully connected to the Holylands. Another would be a 300m extension of the access road needed from King's Bridge to the Queen's Sport Centre car park, disconnected as that may feel.

Instead of the current two-lane road with free parking, something innovative could be installed in its footprint - perhaps a public use running track linking the physical activity of the PEC out onto the waterfront. There's a very nice 400m stretch within the area from Botanic Gardens to the Ormeau Road.

A much-needed underpass would safely speed active travel journeys across the Ormeau Road. What this would create is a linear park connecting Botanic Gardens to the Ormeau Park. Given the location, calling this new stretch the Holylands Embankment Park seems about right. 

On the opposite bank of the river, Annadale Embankment would be retained as a road from the Ormeau Road to Sunnyside Street. With only local access traffic from the Ormeau Road using this stretch, a road diet could slim the carriageway width, allowing more space for people on the riverbank.

Again, the fast link between the Stranmillis Gyratory and Lower Ormeau is cut, but not eliminated - the opposite bank is still open, but slower due to hitting right turn signal controls on the Ormeau Bridge.

This is the most direct link between Stranmillis and Ormeau, and completely inappropriate for that purpose.

Sunnyside Street has the makings of a pleasant village hub with commercial units able to sustain cafes to rival the Ormeau or Stranmillis. But the constant rumble of cars sneaking through this mixed area of dense terracing and leafier streets suppresses street life.

If you think there's no design behind the Stranmillis Gyratory system, Sunnyside Street betrays a cynical favouring of drivers' needs over residential liveability.

Six speed humps are in place, but so gently angled as to barely register to anyone driving through at 30mph. The inclusion of double yellow lines to deter parking on the approach to the Ormeau Road shows this is clearly seen by traffic planners as a critical link across the city.

Blocking off access at the Ormeau Road end would immediately deter through-journeys. This could see the Sunnyside Street exit repurposed as a mini garden which could serve the adjacent Brewbot and Ambrosia with outdoor seating areas.

One blocked street in isolation would likely only displace traffic coming from Ormeau, so additional calming measures on the surrounding streets would be necessary. North of Sunnyside Street only Haywood Avenue has a viable route better than proceeding down to Ormeau Embankment, yet the squeezed street layout and immediate tight right turn give the visual impression of a cul-de-sac.

It's challenging to devise a fool proof system of defeating through-traffic without severely disadvantaging local residents. The first step is to make the whole footway along the eastern side of the upper Ormeau as a continuous pedestrian footway. This works not only as a counter vision to the current highly-swept turns of the current street exits, but is also place-appropriate to the bustling independent shopping district on the Ormeau Road.

Many #currymile minor side-roads should have this 'continuous footway' treatment that priorities #walking & #cycling pic.twitter.com/ynEGVCgztr

— GM Cycling Campaign 🐝🚲 (@GMcycling) September 10, 2016

After that, clever deployment of home zone street treatments, build outs and one-way streets can make through-travel for vehicle drivers awkward, glacially slow and undesirable.

Where Sunnyside Street meets Annadale Embankment, the current layout rewards those wanting a quick short-cut to the west with an almost free-flow exit onto the Stranmillis Gyratory. A fully signal-controlled four-way junction (with a simultaneous green pedestrian and cycling phase) would radically reduce the flow rate from all sides, and tip the risk-reward balance away from through-traffic.

This area already has the most cycling-friendly community in the country, with over 6% of workers commuting by bicycle in 2011. Building upon that foundation and removing unnecessary traffic can only spur more people to do likewise.

And now we reach the crux, the northern side of the Stranmillis Gyratory itself. This bridge, built just before the Titanic, is the eastbound one-way funnel for vehicles.

Once across the bridge, the options for onward travel are too generous. Take an example destination like the major crossroads at Castlereagh Road / Grand Parade - from King's Bridge you can turn left, right or continue straight ahead for three different journey options all within one minute of each other, going via:

• My Lady's Road
• Sunnyside Street
• Annadale Avenue

It's such a porous area you'd almost be mad not to drive. This is where major surgery is needed.

To calm the entire system, we begin by making King’s Bridge two-way with fully signal controlled junctions on either side. No free-flow slips, and an all-green pedestrian and cycling phase to simplify the junction for the most vulnerable users. Additional control adds journey time to the mental map of the surrounding streets, a discouragement in itself.

With two-way traffic continuing across the bridge, underpasses on both riverbanks would free-up many active travellers from the need to interact with vehicle traffic.

The bridge deck itself is ridiculously tight, and feels like a racetrack. Additional boardwalks over either side of the bridge would provide a calmer environment for non-motorised users.

So far, not a great change in terms of restricting traffic. But we're far from done..

It only makes sense if we get serious. If you were wondering why the King's Bridge was being changed to two-way travel, maybe you saw this coming.

To truly address the problems created by the Stranmillis Gyratory system, one bridge needs to close to vehicle traffic. And we favour Governor's Bridge, perhaps against instincts.

It's younger the King's Bridge at around 40 years old. It's capable of carrying heavier vehicles than King's Bridge which has a 7.5 tonne weight restriction. It has a roomier deck and is located close to the Stranmillis roundabout.

But (age apart) all of these aspects make it more attractive to fast and heavy through-traffic. Re-routing traffic along Ridgeway Street provides a handy diversion with greater distance and the visual discouragement of an acute angle turn.

Governor's Bridge itself will of course remain and can be re-purposed as a place for people - somewhere for cycling and walking journeys to continue without the constant danger of fast vehicles, likely with seating on the deck to watch the rowers cutting up and down the river. Perhaps it could even become Belfast's smaller, sensibly-priced answer to London's Garden Bridge mess.

Taking out Governor's Bridge and routing traffic onto King's Bridge isn't worth much if the Stranmillis Embankment between the roundabout and Ridgeway Street still offers a fast route for vehicles, so we propose to remove the road entirely here. Gone.

Instead, a greenway would give a more appropriate link between the forested slopes and river. All traffic between the roundabout and the east bank of the river now must use Ridgeway Street. By lifting the road from roundabout to river, we can create a new mini park.

The existing footprint is dominated by a car park owned and operated by Belfast City Council – 87 free spaces which serve no obvious purpose, given the parking within Stranmillis University grounds and at the businesses opposite.

Although a great opportunity to create a new park space in itself, if you've been watching through the maps to date, it would actually be the southern-most tip of a new 2.5km linear river park stretching (without the need to cross a single road because of new underpasses) from Stranmillis Roundabout, incorporating Botanic Gardens, the new Holylands Embankment Park and across to Ormeau Park.

Which is a wonderful concept for a riverside reclaimed from vehicles.

Where Annadale Avenue meets the Ormeau Road marks the southern event horizon of the Stranmillis Gyratory. This fast carriageway, with a semi-rural feel, draws in traffic from the Saintfield Road and Upper Ravenhill directions, and provides quick access to Forestside shopping complex from west of the Lagan.

Even with the changes laid out so far, getting from Forestside to the Stranmillis roundabout is still attractive - despite the diversion along King's Bridge and Ridgeway Street.

The simplest way to cut through-traffic then is to cut the road itself. There is a natural gap which can be created between developments which grew up from the 1990s onwards, Mornington and Wellington Square.

On the Mornington side, access to the allotments would be retained, but otherwise the road could be lifted to provide new green space or perhaps expand the allotments.

On the Wellington Square side, the split access road would be turned into a large roundabout, doubling up as a terminus for bus services.

The green-light for the construction of the Lagan Gateway will naturally open up improved active travel routes from this section of former road.

Chairman of Lagan Navigation Trust @Sherlocker2 explains the work due to start on first Belfast boat lock in 250yrshttps://t.co/ZsOzrZslIB pic.twitter.com/M5HhsFTV1v

— NICDA (@NICDACOOP) July 25, 2017

With the major road repurposing complete, we turn to the benefits which will accrue to Stranmillis Road itself. No longer part of a through-route, traffic levels will dramatically drop.

One key benefit of this would be the traffic engineering problem-solving rolled out over the years becomes redundant. Most significantly, with only local traffic using the road in ‘rush hour’ the need for a bus lane crawling up the hill towards Stranmillis Village disappears. This can be repurposed into a protected two-way cycleway to link the Village with Lockview.

I've always loved Stranmillis Village but, when you take a step back - it's really horribly car-choked.. pic.twitter.com/XCnLEt9XA5

— NI Greenways (@nigreenways) July 22, 2017

The removal of significant volumes of traffic will probably increase average speeds on the road. The signalised junction at Ridgeway Street will help to calm the approach to the Village section.

Within the Village, the cycleway will need to continue through and on towards the University as Stranmillis and Malone merge again. To achieve this the free parking bays on the west side of the road would be removed, and serious consideration given to the need for any free on-road parking in this picturesque street.

The last key part of the lead-in to the Stranmillis Gyratory is the cut-through from the Lisburn Road to Stranmillis via Chlorine Gardens.

Densely packed with recently-built apartments, older large houses and university buildings, the high volumes of traffic using it daily are harmful and unwelcome. Cutting the road between the vehicle access to Chlorine Mews and the houses beside the new QUB Biological Sciences building would create two woonerfs and a pleasant central pedestrian plaza.

Michael Nugent Ltd Awarded £39m QUB School of Biological Sciences Project!https://t.co/d3Fhus3vTO #MNL #QUB pic.twitter.com/tEkfx2fiI5

— Michael Nugent Ltd (@MICHAELNUGENTLT) July 27, 2016

The five arms of the roundabout become just four with the removal of the road to Governor's Bridge and Stranmillis Embankment. The reduction in traffic means the roundabout can be retained and redesigned to better accommodate the movements of people - pedestrian and cyclist - especially between the Malone and Lockview directions to the new linear park.

The current road layout entering the roundabout from the southern end of Stranmillis Road is a bloated three lanes - two on and one off. Reducing this back to just one on one off, and getting rid of the right hand turning box into Sharman Road, means a fully protected cycleway can be run between the roundabout and Richmond Park.

This would support safe cycling journeys to be made from the Lagan direction to Stranmillis Primary School, another key element of reducing the impact of school run traffic in the area. Similarly, good access across the roundabout onto the new bus-lane-replacement cycleway up to Stranmillis Village offers not just work commuting options but serves Methodist College and Queen's University students.

There it is - a starting point for discussion with 4km of riverside roads cut down to just 2km, replaced by potentially magnificent linear parks and through-traffic all but removed from a whole city district. The Stranmillis Gyratory erased and liveable streets growing up in its place and around its former orbit.

The Ormeau and Stranmillis Embankments don't stand a chance now I've seen Voie Georges Pompidou.. pic.twitter.com/9JDUnegSfo

— NI Greenways (@nigreenways) June 30, 2017

Will this happen tomorrow? No. Will it happen ever? It's doubtful. A massive swathe of the city would be affected by these proposed changes – it would take unprecedented bravery from politicians, planners and the public to even consider it.

But this type of macro-level change is what our city needs to begin to climb out of its chronic congestion and obesity problems.

The underlying framework for movement in the city should be clear - the arterial routes of Belfast and the ring roads are the natural place for longer distance vehicle travel. While the connecting street network continues to remain entirely permeable to vehicles, drivers will continue to use them in great numbers to the detriment of local residents and to the health of the wider population.

Promotion of active travel will only get us so far when the physical space it's expected to flourish in is dominated by vehicles.

This is not to ban vehicles - routes between the Malone and Ormeau arterials are still available in this plan:

• Outer Ring
• Ormeau Embankment / Ridgeway Street
• University Avenue
• Donegall Pass
• Ormeau Avenue (Inner Ring)

Certain short car journeys, especially for those with accessibility and mobility needs, means retaining an appropriate level of vehicle access is important. A turbo roundabout spinning together all types of vehicle journey isn't needed, while the outlying enablers of the worst through-traffic - Ormeau Embankment, Annadale Avenue and Chlorine Gardens - would be rendered inert in this plan.

The primary vehicle usage would be by local residents accessing their properties, and active travel would get promoted through deed, not just word.

In fact, the plan actually has benefits to vehicle travel on the main roads. Ormeau Road loses two side roads with efficiency gains for traffic flow across Ormeau Bridge.

The pedestrian crossing on the northern end of Ormeau Bridge would remain, but would have significantly less usage with a bridge underpass linking the two parks.

Ormeau Bridge at rush hour is Belfast's Bicycle Central (as always) pic.twitter.com/61V9OAafry

— NI Greenways (@nigreenways) September 22, 2016

Traffic turning in and out of Sunnyside Street would no longer cause constant conflict on the upper Ormeau. With no 'strategic' traffic flows along Annadale Avenue, that junction would become far more efficient for Ormeau traffic.

On the Stranmillis Road, local residents would enjoy a drastic reduction in fast traffic flow as they access residential streets. Stress on the Malone Road junction with Stranmillis at Broomhill would be lessened.

Reconnecting communities and our existing public gardens with the riverbanks - as well as the creation of a new linear park - can increase physical activity levels in the surrounding population. Innovative interventions such as urban beaches or public running tracks (or something entirely new created by the folk of Belfast) can fill the space vacated by fast traffic.

The theory goes that reducing road capacity and snipping drivers' short-cuts may displace vehicles elsewhere in the short term, but making public transport and active travel journeys quicker (and safer) on those old routes creates the best incentives to switch travel modes - time and convenience.

This is also just one idea out of many options for the area. It's possible the whole scheme as laid out above is ridiculous (we're prepared to hold our hands up if so) and maybe you have a simpler or better idea. The comments section is open. That's the point - starting an uncomfortable discussion on how we tackle the vehicle use that chokes our city, but that no-one is really planning to eliminate any time soon.

Unhealthy road environments sitting in plain sight have to be tackled, even if we've grown to appreciate their convenience for our driving habits.

How would you kill off the Stranmillis Gyratory?

Footnote

Had motorway planning in the 1960s been seen through, the situation here might have been infinitely worse. The M4 motorway from Carryduff was originally planned to interchange with the Belfast Urban Motorway inner ring, ploughing its way along the banks of the Lagan.

In a revised 1969 plan shown below, having being deemed too destructive, the M4 was to terminate at Annadale Embankment - which would have mainlined strategic traffic right onto the Stranmillis Gyratory.

Today's remaining wide roads on both banks are a little contemporary echo of that half-a-century-old mode of thinking about urban travel priorities.

Picture courtesy of Wesley Johnston (NI Roads) / DfI

Some other copyright images (as marked) reused under Creative Commons Licence from the cherished lens of Albert Bridge.

Inclusion of all of these images does not indicate endorsement of the article.

Update 

01 August 2017

By popular demand (in the comments section) I've crafted "before and after" full scheme maps in PNG formats for your viewing and downloading pleasure!

Full map of existing road network on scheme footprint (PNG, 503K)

Full map of proposed alterations to road network (PNG, 503K)