Why do British Highway Engineers Hate Free-Flow Junctions?

[Mordac]

The continental country whose roads I'm more familiar with is Portugal, but also done my share of driving in the US and Canada. As per the thread title, it seems that only under duress will UK highway engineers vary the old "bung a roundabout on top" approach. Junctions that would get a free-flow in any other country I'm familiar with, like Stivichal interchange in Coventry, or Kegworth interchange near Nottingham in the M1, seem to get saddled with roundabouts creating a bottleneck in what would otherwise be an entirely free-flow dual carriageway. With all the current concern about emissions, surely they should know that making cars slow down and stop, and then accelerate again, is much worse than them remaining at a more or less constant speed.

I read a while ago that there are only three clover-leaf interchanges in the UK (one of them is near me in Redditch). Again, these are ubiquitous pretty much everywhere else. I really don't understand this antipathy. Can someone shed a light?

 

[Dai Corner]

The continental country whose roads I'm more familiar with is Portugal, but also done my share of driving in the US and Canada. As per the thread title, it seems that only under duress will UK highway engineers vary the old "bung a roundabout on top" approach. Junctions that would get a free-flow in any other country I'm familiar with, like Stivichal interchange in Coventry, or Kegworth interchange near Nottingham in the M1, seem to get saddled with roundabouts creating a bottleneck in what would otherwise be an entirely free-flow dual carriageway. With all the current concern about emissions, surely they should know that making cars slow down and stop, and then accelerate again, is much worse than them remaining at a more or less constant speed.

I read a while ago that there are only three clover-leaf interchanges in the UK (one of them is near me in Redditch). Again, these are ubiquitous pretty much everywhere else. I really don't understand this antipathy. Can someone shed a light?

Cost and the amount of land required.

 

[Bletchleyite]

Cost and the amount of land required.

And roundabouts are very simple, easy to use and all destinations can easily be chosen. I would say I prefer them myself.

 

[Ediswan]

Cloverleaf junctions are only free-flowing when traffic is light. If the junction gets too busy, there is an almighty bunfight where traffic entering from one loop has to cross paths with traffic seeking to exit usng the immediately following loop.

Don't get me started on M1 J15. It will probably be fine when fully signed. At present, not pleasant.

 

[swt_passenger]

Cloverleaf junctions are only free-flowing when traffic is light. If the junction gets too busy, there is an almighty bunfight where traffic entering from one loop has to cross paths with traffic seeking to exit usng the immediately following loop.

I suppose that’s exactly why so few true cloverleafs were ever built in this country. By the time they were being considered experience from other countries led to them being discounted?

 

[Snow1964]

We do things on the cheap, build a roundabout, then few years later scatter some traffic lights around it, as it doesn’t work as a roundabout

Seems that any roundabout with more than two lanes around it, is now infected with traffic lights, which does rather imply that big multi-lane roundabouts have been a design failure.

Another bugbear of mine is 4way traffic lights with single lane approaches, making all traffic for right, left and straight ahead queue in one lane which means each phase is slow.

 

[Bletchleyite]

Cloverleaf junctions are only free-flowing when traffic is light. If the junction gets too busy, there is an almighty bunfight where traffic entering from one loop has to cross paths with traffic seeking to exit usng the immediately following loop.

Don't get me started on M1 J15. It will probably be fine when fully signed. At present, not pleasant.

Isn't that just a slightly convoluted and presently badly signed roundabout?

 

[lachlan]

With all the current concern about emissions, surely they should know that making cars slow down and stop, and then accelerate again, is much worse than them remaining at a more or less constant speed.

On the other hand, more road capacity induces demand, which results in increased emissions. Stopping and starting traffic will also be less of a concern with electric cars

 

[stuu]

Stupid short-sightedness. We do things to save money, and then have to come back in 20 years and fix the problem at vast expense, whilst costing the economy millions in lost time. See A14-M6 amongst many others.

 

[Snow1964]

Stupid short-sightedness. We do things to save money, and then have to come back in 20 years and fix the problem at vast expense, whilst costing the economy millions in lost time. See A14-M6 amongst many others.

As an example, the roundabout linking M25 (junction 10) and A3 is due to be upgraded at a cost of £317 million, simply because didn’t put a free flow junction in initially, even though both roads had 3 lanes each way

Start date Summer 2022
End date Summer 2025
Cost £317 million

Our proposals will improve connections, smooth traffic flows and create safer journeys for all.

We are proposing four new slip roads for the M25 junction 10 roundabout, meaning traffic turning left can pass through the junction easily.

The M25 will increase from three to four lanes through the junction, with the A3 also becoming four lanes either side of the junction 10 roundabout. There will also be improvements to the A245, Sevenhills junction and Painshill roundabout.

Around the junction, there will be new and safer routes for pedestrians, cyclists and horse riders. Improvements will also be made to the local environment and wildlife.

M25 junction 10 - National Highways

We're restoring heathland and upgrading the junction with the A3 Wisley interchange and to reduce congestion, improve safety and create more reliable journeys.

nationalhighways.co.uk

 

[stuu]

As an example, the roundabout linking M25 (junction 10) and A3 is due to be upgraded at a cost of £317 million, simply because didn’t put a free flow junction in initially, even though both roads had 3 lanes each way

M25 junction 10 - National Highways

We're restoring heathland and upgrading the junction with the A3 Wisley interchange and to reduce congestion, improve safety and create more reliable journeys.

nationalhighways.co.uk

Not a great example, as HE are replacing a congested roundabout with a bigger roundabout and spending £300m to do so. That's fairly close to insanity

 

[MotCO]

Whilst not prone to long queues, I can think of two motorway spur junctions which could benefit from freeflow lanes - Junction 4 of M25, and the Gatwick turnoff on the M23 (but this may now have been fixed). At Junction 4 of M25, traffic from Sevenoaks could flow directly on to the spur road if the traffic from Swanley was restricted to a single lane, which is not too problematic.

And yes, traffic lights are not always the answer. On a slight tangent, the traffic lights at Blindley Heath on the A22 (the Lingfield turnoff) default to red when no traffic is around. When traffic is detected, the lights turn green, so you can have the situation where the lights turn to red then straight back to green. It seems to work well - does this happen anywhere else?

 

[Hadders]

Don't get me started on M1 J15. It will probably be fine when fully signed. At present, not pleasant.

Isn't that just a slightly convoluted and presently badly signed roundabout?

M1 J15 is a right mess at the moment. It'll probably flow better once all the work's finished but it's a hige, huge project that's already been going on for months.

It's very easy to be travelling on the A45 from Northampton, intending to join the M1 south and to completely miss the lane you need to be in.

 

[D365]

We do things on the cheap, build a roundabout, then few years later scatter some traffic lights around it, as it doesn’t work as a roundabout

Seems that any roundabout with more than two lanes around it, is now infected with traffic lights, which does rather imply that big multi-lane roundabouts have been a design failure.

It's the curse of piecemeal upgrades. Black Cat (and the former Catthorpe) says hello.

... or Kegworth interchange near Nottingham in the M1...

In fairness - is there really the land to do more?

 

[Ediswan]

It's very easy to be travelling on the A45 from Northampton, intending to join the M1 south and to completely miss the lane you need to be in.

Exactly. I missed the lane split for M1(S). Then 'found' the lane split for M1(N). On the bright side, not far to J15A.

 

[Ken H]

They have neqrly finished a new road east of leeds from the M1 to Shadwell so the ring road bypasses Crossgates and Seacroft. But no grade separated junctions. So loads of roundabouts. And there is a roundabout on this new road where it crosses the A64 York Road.
That will snarl up pretty quickly.

 

[Bald Rick]

As an example, the roundabout linking M25 (junction 10) and A3 is due to be upgraded at a cost of £317 million, simply because didn’t put a free flow junction in initially, even though both roads had 3 lanes each way

to be fair that’s been open 39 years (I well remember opening day!)

 

[telstarbox]

We do things on the cheap, build a roundabout, then few years later scatter some traffic lights around it, as it doesn’t work as a roundabout

Seems that any roundabout with more than two lanes around it, is now infected with traffic lights, which does rather imply that big multi-lane roundabouts have been a design failure.

Another bugbear of mine is 4way traffic lights with single lane approaches, making all traffic for right, left and straight ahead queue in one lane which means each phase is slow.

Can you give an example of the last one please?

 

[DelW]

It's certainly not true that Britain resisted free-flow junctions - the grand-daddy of them all, Gravelly Hill on the M6 in north Birmingham, was built in the late 1960s. (I watched it being built from my bus going to and from school).

Cloverleafs indeed weren't popular, they take up far too much room unless the inner loops are very tight, and as mentioned don't cope well with high volumes. The preferred British motorway to motorway freeflow junction is the four-level stack, as at Almondsbury (M4/M5) and Merstham (M25/M23), or the looped slip roads as at Lyne (M25/M3). Both of these designs cope with high traffic flows far better than cloverleafs and take up less space, although they do need more bridges.

It's possible to retrofit freeflow links to a roundabout junction - nearly 40 years ago I was building bridges (over the M20 and under the M25) to carry the new link between the M20 and M25 at Swanley as an addition to the original junction. However, the M25 was then still incomplete and carrying much less traffic than now - I wouldn't want to try to put contraflows onto those roads in today's traffic.

 

[Mordac]

Cost and the amount of land required.

Let's take a particularly ridiculous example that I forgot yesterday. M5 J8, roundabout junction between two motorways (!) in the middle of nowhere. What would be the incremental cost of putting a Trumpet interchange in there instead?

And roundabouts are very simple, easy to use and all destinations can easily be chosen. I would say I prefer them myself.

Roundabouts are easy to use if they are empty, just like any other junction. Once they get congested, getting the guy in the Land Rover to let you change into the right lane after you've made a mistake isn't particularly easy either. I use the mother of all free flow junctions, Spaghetti Junction, fairly often and I find it very easy to use.

 

[stuu]

Let's take a particularly ridiculous example that I forgot yesterday. M5 J8, roundabout junction between two motorways (!) in the middle of nowhere. What would be the incremental cost of putting a Trumpet interchange in there instead?

It was a trumpet originally, lorries kept falling over so it was removed when the M5 was widened in the late 80s

 

[Dai Corner]

Let's take a particularly ridiculous example that I forgot yesterday. M5 J8, roundabout junction between two motorways (!) in the middle of nowhere. What would be the incremental cost of putting a Trumpet interchange in there instead?


Roundabouts are easy to use if they are empty, just like any other junction. Once they get congested, getting the guy in the Land Rover to let you change into the right lane after you've made a mistake isn't particularly easy either. I use the mother of all free flow junctions, Spaghetti Junction, fairly often and I find it very easy to use.

The cost would almost certainly exceed the benefit. The M50 carries very little traffic.

 

[Julia]

Cloverleaf junctions are only free-flowing when traffic is light. If the junction gets too busy, there is an almighty bunfight where traffic entering from one loop has to cross paths with traffic seeking to exit usng the immediately following loop.

Although not a full cloverleaf, the A14/M11 at Girton was notorious for jamming up with two solid streams of traffic (A14 westbound, which is a "right turn" not the straight route!, versus M11 northbound -> A14 eastbound) having to physically pass through each other on the bridge over the A14/A428 carriageway. Accidents galore.

 

[ABB125]

Excellent thread choice!

Stupid short-sightedness. We do things to save money, and then have to come back in 20 years and fix the problem at vast expense, whilst costing the economy millions in lost time. See A14-M6 amongst many others

Precisely this - penny-pinching by (usually) the Treasury is a big problem.
Gold-plated standards don't help though. Neither does the issue of consultants deliberately creating future work for themselves, not the seeming lack of appetite by road authorities (local or national) to challenge what their consultants tell them (and housing developers...)

As an example, the roundabout linking M25 (junction 10) and A3 is due to be upgraded at a cost of £317 million, simply because didn’t put a free flow junction in initially, even though both roads had 3 lanes each way

"Upgrade" isn't really the word I'd use. "Complete utter waste of money" is more appropriate

In fairness - is there really the land to do more?

There likely was 20-30 years ago when an overall masterplan for the current mess that is the Kegworth junction complex should have been created

Let's take a particularly ridiculous example that I forgot yesterday. M5 J8, roundabout junction between two motorways (!) in the middle of nowhere. What would be the incremental cost of putting a Trumpet interchange in there instead?

It used to be a trumpet (you can see where the loop used to be on the eastern side of the M5). When the M5 was widened* (due to, surprise surprise, short sightedness of building the bit between Birmingham (J3 maybe?) and J8 as two lanes; south of J8 has always been three) it was changed to a roundabout (Strensham Services were also rebuilt at this time).

*Of course, had the correct decision been made to build a motorway between Strensham and the M42/M40 junction, as nearly happened twice, instead of widening, we'd have ended up with a much more interesting (and free flow) design (I'll post some links later).

Occasionally, National Highways(' consultants) can come up with some kind of free-flow junction. Look at the spaghetti fests that are planned where the lower Thames crossing meets the A13 and M2. Although on that subject, we come to the stupidity that the LTC will be a restricted A road (like the new A14 at Huntington) rather than a motorway).

As for roundabouts, small, simple ones are fine. Enormous, multi-lane signalled ones are an abomination that screams "we don't know how to do anything other than add more lanes and traffic lights and hope for the best".


More posting will follow this evening once I'm back home...

Although not a full cloverleaf, the A14/M11 at Girton was notorious for jamming up with two solid streams of traffic (A14 westbound, which is a "right turn" not the straight route!, versus M11 northbound -> A14 eastbound) having to physically pass through each other on the bridge over the A14/A428 carriageway. Accidents galore.

Cloverleaves can work fine if there's comparatively little turning traffic. "Improved" cloverleaves are better, as all weaving occurs on a separe carriageway from the straight through route (such as here:
E35
https://maps.app.goo.gl/wBhYCiKELvhsZxdr6). However for high turning volumes (such as the A14 example you mentioned) the don't really work due to the weaving (though you can "solve" that problem by grade-separating the loops, such as here in Frankfurt: A3

https://maps.app.goo.gl/WqPZZ9FEDW81rZGi8).

The good thing about cloverleaves is that they're really cheap (a single bridge is needed, less than a roundabout!) and can be upgraded by "opening out" the loops fairly easily (see here: A7

https://maps.app.goo.gl/wAqabiR1C2pB9RVQ8)

 

[Mordac]

I suppose that’s exactly why so few true cloverleafs were ever built in this country. By the time they were being considered experience from other countries led to them being discounted?

Other countries haven't stopped building them, though.

We do things on the cheap, build a roundabout, then few years later scatter some traffic lights around it, as it doesn’t work as a roundabout

Seems that any roundabout with more than two lanes around it, is now infected with traffic lights, which does rather imply that big multi-lane roundabouts have been a design failure.

Another bugbear of mine is 4way traffic lights with single lane approaches, making all traffic for right, left and straight ahead queue in one lane which means each phase is slow.

Stivichall Interchange in Coventry is almost the paradigm case for that.

Not a great example, as HE are replacing a congested roundabout with a bigger roundabout and spending £300m to do so. That's fairly close to insanity

Again, same as Stivichall.

It's the curse of piecemeal upgrades. Black Cat (and the former Catthorpe) says hello.


In fairness - is there really the land to do more?

Perhaps not now, although at the very least it looks like there's the space to let A50 to A453 traffic flow freely. But as mentioned above, a masterplan for this mess should have been made a long time ago.

It's certainly not true that Britain resisted free-flow junctions - the grand-daddy of them all, Gravelly Hill on the M6 in north Birmingham, was built in the late 1960s. (I watched it being built from my bus going to and from school).

Cloverleafs indeed weren't popular, they take up far too much room unless the inner loops are very tight, and as mentioned don't cope well with high volumes. The preferred British motorway to motorway freeflow junction is the four-level stack, as at Almondsbury (M4/M5) and Merstham (M25/M23), or the looped slip roads as at Lyne (M25/M3). Both of these designs cope with high traffic flows far better than cloverleafs and take up less space, although they do need more bridges.

It's possible to retrofit freeflow links to a roundabout junction - nearly 40 years ago I was building bridges (over the M20 and under the M25) to carry the new link between the M20 and M25 at Swanley as an addition to the original junction. However, the M25 was then still incomplete and carrying much less traffic than now - I wouldn't want to try to put contraflows onto those roads in today's traffic.

They will build free-flows if there is no possible alternative, but they need to be dragged kicking and screaming into it, unlike almost everywhere else.

The four-level stack is definitely better than a cloverleaf at handing extremely high levels of traffic, but that's the point, they'll only build free-flows when the levels of traffic are extremely high, whereas in most other countries the idea of connecting two motorways or expressways with roundabounds would not even be contemplated.

It was a trumpet originally, lorries kept falling over so it was removed when the M5 was widened in the late 80s

That is actually insane. Why not just change the curve radius??

Stupid short-sightedness. We do things to save money, and then have to come back in 20 years and fix the problem at vast expense, whilst costing the economy millions in lost time. See A14-M6 amongst many others.

And of course, they only did that once they made sure that traffic couldn't go from the A14 into the M6 Northbound, which is absolutely nuts.

 

[ABB125]

They will build free-flows if there is no possible alternative, but they need to be dragged kicking and screaming into it, unlike almost everywhere else.

Case point being Lofthouse (M1/M62), where even National Highways have basically admitted that anything other than a 4-level stack won't work.

That is actually insane. Why not just change the curve radius??

To be honest, this is one junction where traffic volumes don't really justify free-flow. I've never been held up there, and I think it only grinds to a halt of the M5 is shut somewhere are all the traffic is heading down the M50 instead. I'm slightly surprised it hasn't been "upgraded" with traffic lights though...
[cynic] the only northbound services have become a maintenance depot, and access to maintenance depots trumps everything (even though they could just use the original local road network access from the services in order to go in either direction) [/cynic]

 

[stuu]

The four-level stack is definitely better than a cloverleaf at handing extremely high levels of traffic, but that's the point, they'll only build free-flows when the levels of traffic are extremely high, whereas in most other countries the idea of connecting two motorways or expressways with roundabounds would not even be contemplated.

What is particularly stupid is that they build new roads with grade-separation for the minor junctions, and then use roundabouts for the major connections - even worse the planned A358 dualling in Somerset will achieve the feat of having two roundabouts in close succession where it meets the M5

I believe that, apart from finances, part of the problem is that the design code for new roads is very prescriptive and limiting in what can be done, so pragmatic ideas like tighter loops for freeflow are very difficult to justify

 

[jon0844]

Whilst not prone to long queues, I can think of two motorway spur junctions which could benefit from freeflow lanes - Junction 4 of M25, and the Gatwick turnoff on the M23 (but this may now have been fixed). At Junction 4 of M25, traffic from Sevenoaks could flow directly on to the spur road if the traffic from Swanley was restricted to a single lane, which is not too problematic.

And yes, traffic lights are not always the answer. On a slight tangent, the traffic lights at Blindley Heath on the A22 (the Lingfield turnoff) default to red when no traffic is around. When traffic is detected, the lights turn green, so you can have the situation where the lights turn to red then straight back to green. It seems to work well - does this happen anywhere else?

Traffic lights in Sweden in the 80s?

 

[Bletchleyite]

I believe that, apart from finances, part of the problem is that the design code for new roads is very prescriptive and limiting in what can be done, so pragmatic ideas like tighter loops for freeflow are very difficult to justify

I'm a driver and a taxpayer. As a driver I have no issue with roundabouts, and indeed prefer them to overly tight cloverleafs. As a taxpayer I want good value for money.

I therefore see no issue with the status quo.

 

[Mordac]

I'm a driver and a taxpayer. As a driver I have no issue with roundabouts, and indeed prefer them to overly tight cloverleafs. As a taxpayer I want good value for money.

I therefore see no issue with the status quo.

But do you care that people will endure needless congestion in the future until the next generation decides to stump of thrice as much as what it would have cost to build things to a proper standard in the first place?