Collector/Distributor Roads and Express Lanes

[Chris5156]

There are a couple of examples in France - you're probably thinking of when the A1 was widened in the 1990s here: https://maps.app.goo.gl/Uqgc5onncs1yBnmg8 (sortie 17.1 came much later than the widening).

Thanks, that's the one, or at least the kind of thing I had in mind. To me those aren't C/D lanes, though I guess Peter would say they are, given what he says about a hypothetical M20 J6 without J5.

Of course that's a C/D road! By staying on the express lanes (keep left), you don't have to deal with a diverge and a merge associated with sortie 17.1. That's precisely what C/Ds are for - to stretch out the distance between junctions. It removes diverge/merge interactions from the mainline.

No, I call nonsense on that. It's not collecting or distributing anything. It's just a place where the mainline lanes split around an obstacle.

There are two unusual points about this French one -
1. It bypasses only one junction (same as M60 and M20).

I disagree, it bypasses no junctions!

[MotorwayGuy]

Thanks, that's the one, or at least the kind of thing I had in mind. To me those aren't C/D lanes, though I guess Peter would say they are, given what he says about a hypothetical M20 J6 without J5.

Of course that's a C/D road! By staying on the express lanes (keep left), you don't have to deal with a diverge and a merge associated with sortie 17.1. That's precisely what C/Ds are for - to stretch out the distance between junctions. It removes diverge/merge interactions from the mainline.

No, I call nonsense on that. It's not collecting or distributing anything. It's just a place where the mainline lanes split around an obstacle.

There are two unusual points about this French one -
1. It bypasses only one junction (same as M60 and M20).

I disagree, it bypasses no junctions!

Looks like it was originally a D2, and it splits to go over this extended bridge. The offside lane on the A13 splits over this flyover for some reason. I'm guessing it's something to do with the structure because it ends just before the junction, for safety reasons it would make sense for it to continue up to the diverge.

[Chris5156]

The offside lane on the A13 splits over this flyover for some reason. I'm guessing it's something to do with the structure because it ends just before the junction, for safety reasons it would make sense for it to continue up to the diverge.

It was originally a pair of flyovers forming a two-lane dual carriageway, that was later widened by putting another two-lane structure alongside. Seen from the underneath it's obvious. To get three lanes in each direction, the middle structure now carries one lane each way. Because they are physically separate structures you can't just join them up into conventional three-lane carriageways.

You're right that the carriageway divider could continue further to discourage dangerous lane changing to reach the diverge just to the west. In many ways the divider could be better handled - longer approach markings and better signage would avoid lots of near-misses where the barrier starts. You could easily paint a solid line for a reasonable distance beforehand to discourage lane changes; instead there's a very short and very narrow painted taper just before the barrier begins.

[freebrickproductions]

The offside lane on the A13 splits over this flyover for some reason. I'm guessing it's something to do with the structure because it ends just before the junction, for safety reasons it would make sense for it to continue up to the diverge.

It was originally a pair of flyovers forming a two-lane dual carriageway, that was later widened by putting another two-lane structure alongside. Seen from the underneath it's obvious. To get three lanes in each direction, the middle structure now carries one lane each way. Because they are physically separate structures you can't just join them up into conventional three-lane carriageways.

(Emphasis mine)

I'd imagine it'd probably come down to how the new bridge was built for the widening, and maybe the expected traffic load. I know here in Alabama, ALDOT was able to widen an existing bridge on US 31/AL 22 over I-65 in Clanton by building what appears to essentially be a new, one-lane bridge that's tied into the pre-existing one, to allow for the US 31/AL 22 bridge to go from two lanes to three over I-65:
https://www.google.com/maps/@32.8020035 ... ?entry=ttu
The bridge even now has two different designs of railing on the sides of it as a result:
https://www.google.com/maps/@32.8018295 ... ?entry=ttu

I suspect something similar was done to the bridges along I-65 between Prattville/Millbrook and Montgomery, though I can't easily prove it without finding a way to view the supports of them. The concrete on the bridges is a lot smoother (and seems to be a lot newer) on the left/inside lanes on both sides though, compared to the middle and outside lanes:
https://www.google.com/maps/@32.4024864 ... ?entry=ttu

[Peter Freeman]

Does the use of a parallel road (A1(M) in Cambridgeshire and Yorkshire or A /M74) count as an alternative to C/D lanes?

You can use them as alternatives, but they're not C/Ds. They're local access roads (LAR). Even though they may eventually return to the motorway that they left, they usually include, or pass through, local junctions of their own. Virtually all of them are 'leftovers', after being bypassed by offline upgrades.

C/Ds leave and re-join the express mainline via diverge/merge or lane drop/add. LARs leave and re-join motorways at GSJs.

C/Ds are motorway-standard, LARs are not.

[Peter Freeman]

... I don't really see C/D lanes as the gold standard they're sometimes held up as, as the additional connections to the mainline are inimical to traffic flow, creating extra weaving zones, decision points and merges, compared to the "long sliproad" design that only connects at one end like the M8 at Plantation ...

1. "I don't really see C/D lanes as the gold standard" - nor do I (which is also a view that I've previously expressed - excuse the pun). In most cases, I'd opt simply for a wider Dx. Almost always, where you could have a C/D formation of 2-3-3-2 or 3-2-2-3, a D6 would fit instead. D6 provides the same weaving relief (sort-of anyway - stay away from the left-most lanes), and (bonus) carries more traffic; and is simpler and more comprehensible (everyone sees where to go, how to exit, etc).

2. "... compared to the "long sliproad" design ..." - this makes long slips and C/Ds sound interchangeable, ie. a design conundrum might be soluble by either one or the other. This will at times be the case. The usual conundrum is closely spaced junctions resulting in short weaving lengths on the mainline. Sometimes that needs a C/D (or a wider carriageway?), sometimes it needs a braid or two (which may imply a long sliproad). In only some cases, either may be used.

[Peter Freeman]

The A299 Thanet Way heading westbound is another "close, but not quite" example.

Taking the exit signposted for Herne Bay, you pass not one, not two, but three junctions - all of which can't be reached from the mainline.

The sliproad starts here:

https://www.google.co.uk/maps/@51.36045 ... ?entry=ttu

...and almost immediately there's a "map style" sign for the first junction:

https://maps.app.goo.gl/mLR96WoVNP8imAV9A

The next junction is just signed as a minor one, and after that is a business park junction:

https://maps.app.goo.gl/nKB331fNxMTst1qU6

All the while, the road (2 lanes) runs parallel to the A299 mainline. It then finally reaches the Herne Bay junction, which is in the form of a roundabout. You can then go straight over the roundabout to rejoin the mainline of the A299.

So - although it's close to a C/D lane (and would be, if that roundabout wasn't there), it's instead a long, wide sliproad with several junctions - still pretty unusual, I'd have thought.

Yes, that's a good example of 'almost'. Its reason for existing is probably the high ground south of the A299 there (it can be noticed in streetview, and there are names Hillcroft and Hillbrow). The slightly elevated pseudo-C/D road allows the A299 to access that tricky area.

I think we could find a few examples of such near-misses. The M8 between Bailliston and Newhouse (pointed out by Wrinkly and commented on by Jackal) is close, and there's more on M8 a bit further in towards Glasgow.

I've always thought that all the M8 interchanges between J14 and J23 are pretty good.

[solocle]

The A299 Thanet Way heading westbound is another "close, but not quite" example.

Taking the exit signposted for Herne Bay, you pass not one, not two, but three junctions - all of which can't be reached from the mainline.

The sliproad starts here:

https://www.google.co.uk/maps/@51.36045 ... ?entry=ttu

...and almost immediately there's a "map style" sign for the first junction:

https://maps.app.goo.gl/mLR96WoVNP8imAV9A

The next junction is just signed as a minor one, and after that is a business park junction:

https://maps.app.goo.gl/nKB331fNxMTst1qU6

All the while, the road (2 lanes) runs parallel to the A299 mainline. It then finally reaches the Herne Bay junction, which is in the form of a roundabout. You can then go straight over the roundabout to rejoin the mainline of the A299.

So - although it's close to a C/D lane (and would be, if that roundabout wasn't there), it's instead a long, wide sliproad with several junctions - still pretty unusual, I'd have thought.

Yes, that's a good example of 'almost'. Its reason for existing is probably the high ground south of the A299 there (it can be noticed in streetview, and there are names Hillcroft and Hillbrow). The slightly elevated pseudo-C/D road allows the A299 to access that tricky area.

I think we could find a few examples of such near-misses. The M8 between Bailliston and Newhouse (pointed out by Wrinkly and commented on by Jackal) is close, and there's more on M8 a bit further in towards Glasgow.

I've always thought that all the M8 interchanges between J14 and J23 are pretty good.

A11 in Norwich has an elevated stretch, although it's more of a frontage road.

[FtoE]

Whatever has been done to the A9 / A85 junction in Perth (and I still can’t get my head round it) feels a bit like a very short collector/express system - but is really just very long slip roads.

[Peter Freeman]

Of course that's a C/D road!

No, I call nonsense on that. It's not collecting or distributing anything. It's just a place where the mainline lanes split around an obstacle.

OK, I concede on this one.

[Herned]

1. "I don't really see C/D lanes as the gold standard" - nor do I (which is also a view that I've previously expressed - excuse the pun). In most cases, I'd opt simply for a wider Dx. Almost always, where you could have a C/D formation of 2-3-3-2 or 3-2-2-3, a D6 would fit instead. D6 provides the same weaving relief (sort-of anyway - stay away from the left-most lanes), and (bonus) carries more traffic; and is simpler and more comprehensible (everyone sees where to go, how to exit, etc).

I disagree. Where they are done properly they work well, there just aren't very many in the UK. An example I drove along a couple of years ago was the northern bypass of Bologna, which has proper separation for the full distance around the northern side of the city, with ~1km between junctions on the distributor road and only a couple of intermediate connections into the express lanes. I find it very hard to imagine that it would flow was well being a D5 rather than 2+3. But I would be interested to see any traffic modelling of the different scenarios

[Peter Freeman]

1. "I don't really see C/D lanes as the gold standard" - nor do I (which is also a view that I've previously expressed - excuse the pun). In most cases, I'd opt simply for a wider Dx. Almost always, where you could have a C/D formation of 2-3-3-2 or 3-2-2-3, a D6 would fit instead. D6 provides the same weaving relief (sort-of anyway - stay away from the left-most lanes), and (bonus) carries more traffic; and is simpler and more comprehensible (everyone sees where to go, how to exit, etc).

I disagree. Where they are done properly they work well, there just aren't very many in the UK. An example I drove along a couple of years ago was the northern bypass of Bologna, which has proper separation for the full distance around the northern side of the city, with ~1km between junctions on the distributor road and only a couple of intermediate connections into the express lanes. I find it very hard to imagine that it would flow was well being a D5 rather than 2+3. But I would be interested to see any traffic modelling of the different scenarios

I have no evidence to back up my opinion, and anyway all these things depend on the particular case. BTW, I compared 2-3-3-2 with D6, not D5.

In the general case of multiple carriageways versus a single wider carriageway with the same number of lanes, I believe that single will perform better. For example, 4-4 beats 2-2-2-2 in capacity. My reasoning -

The 4-4 would max out when every lane maxes out, with a similar volume (let's say 2000 pcu/hr) in each lane. There is nothing to prevent this distribution, as drivers change lanes to fill gaps. For the 2-2-2-2 to operate with each lane at that flow, the inner and outer carriageways must each carry exactly half of the total. Otherwise, there is under-utilised road surface. This equal occupation is unlikely to occur naturally. For example, when building the road, how can you accurately predict the ratio of carriageway/carriageway loading? And that may vary with time - for example, new junctions introduced, nearby road adjustments etc.

So that's my general reasoning. It may or may not apply in particular circumstances.

The Bologna half-ring-road that you cite is truly impressive (on G.Maps/G.Earth). It has continuous C/Ds all the way around, as you say, and very frequent junctions. The cross-section is 2-3-3-2. The outside carriageways have a shoulder, the inner ones don't (ie. ALR). Melbourne, AU, is currently commissioning a new 6km length of C/D (Westgate Freeway). It is 3-3-3-3, but it's the inner carriageways that have a HS, outers are ALR. Food for thought.

I see from G.Maps traffic layer (and it's discernable on some streetview shots) that the motorway does have considerable peak congestion, but only on the outer carriageways. I suppose this means that the design is working, though one wonders whether the format should perhaps be 3-2-2-3 instead of 2-3-3-2. I notice that there are no voluntary transfers express-to-C/D and vice versa. Also, there are only 2 or 3 places (mainly major radials, as you mentioned) where a junction offers a choice: interchange with either express or C/D . These interchanges are very well designed, but I wonder whether this, combined with the lack of rollovers, discourages express usage. Drivers will be wary of becoming trapped on the inner carriageway and missing their exit.

All-in-all though, a most impressive road - thanks for bringing it up. I do wonder though, how a D5 or D6 version would look and work ...

[Herned]

My reasoning -

The 4-4 would max out when every lane maxes out, with a similar volume (let's say 2000 pcu/hr) in each lane. There is nothing to prevent this distribution, as drivers change lanes to fill gaps. For the 2-2-2-2 to operate with each lane at that flow, the inner and outer carriageways must each carry exactly half of the total. Otherwise, there is under-utilised road surface. This equal occupation is unlikely to occur naturally. For example, when building the road, how can you accurately predict the ratio of carriageway/carriageway loading? And that may vary with time - for example, new junctions introduced, nearby road adjustments etc.

I can see the logic, but on the other hand, the influence of traffic joining and weaving is going to reduce the throughput, potentially significantly. With separated carriageways you don't get that, so the express lanes are likely to work nearer their maximum capacity more of the time. But it's obviously a complicated system to model

I see from G.Maps traffic layer (and it's discernable on some streetview shots) that the motorway does have considerable peak congestion, but only on the outer carriageways. I suppose this means that the design is working, though one wonders whether the format should perhaps be 3-2-2-3 instead of 2-3-3-2. I notice that there are no voluntary transfers express-to-C/D and vice versa. Also, there are only 2 or 3 places (mainly major radials, as you mentioned) where a junction offers a choice: interchange with either express or C/D . These interchanges are very well designed, but I wonder whether this, combined with the lack of rollovers, discourages express usage. Drivers will be wary of becoming trapped on the inner carriageway and missing their exit.

All-in-all though, a most impressive road - thanks for bringing it up. I do wonder though, how a D5 or D6 version would look and work ...

I do agree that the Bologna system, and indeed probably any other around a major city, would be better off giving the capacity to the local lanes rather than the through lanes, there will pretty much always be more local journeys than longer distance.

I suppose unless you built another city with an identical layout but had a D5 ring road instead, it will never be possible to prove either way

[Owain]

I am kicking myself for not thinking of the A14 around Bologna...

I'll see that and raise you the truly remarkable Sopraelevata Aldo Moro through the heart of Genoa.

I've driven on it, and under it. It plays havoc with your SatNav!

[Peter Freeman]

I can see the logic, but on the other hand, the influence of traffic joining and weaving is going to reduce the throughput, potentially significantly. With separated carriageways you don't get that, so the express lanes are likely to work nearer their maximum capacity more of the time. But it's obviously a complicated system to model.

Exactly. That weaving influence is the argument for C/D. I think uninterrupted express lanes do work at, or much nearer to, theoretical maximum capacity, so the C/D argument certainly has a case. My counter-argument is that on a multilane un-divided carriageway, each lane-step to the right brings you closer to a lane that is uninterrupted enough to reach max capacity. At which step (Lane 4, 5 or 6?) is hard to say!

I suppose unless you built another city with an identical layout but had a D5 ring road instead, it will never be possible to prove either way.

Indeed we'll never know.

BTW, it occured to me that the dividing barrier could be simply shifted over by one lane, converting 2-3-3-2 to 3-2-2-3. At first thought it would seem to be easy engineering, low-ish cost. However, looking in detail, that would be impossible almost everywhere on this road.

[Herned]

Exactly. That weaving influence is the argument for C/D. I think uninterrupted express lanes do work at, or much nearer to, theoretical maximum capacity, so the C/D argument certainly has a case. My counter-argument is that on a multilane un-divided carriageway, each lane-step to the right brings you closer to a lane that is uninterrupted enough to reach max capacity. At which step (Lane 4, 5 or 6?) is hard to say!

The other factor, at least in my experience, is that once you get above three lanes, a lot of drivers haven't the faintest idea which lane they should be in. Which, outside of peaks where every lane is full, serves to reduce the real-life capacity of the road. I drove along the M4 into London recently and for large distances the inner 2 lanes were virtually empty. I find that is common on 4 lane sections. That's an education problem, but no one wants to tackle it

[ManomayLR]

Exactly. That weaving influence is the argument for C/D. I think uninterrupted express lanes do work at, or much nearer to, theoretical maximum capacity, so the C/D argument certainly has a case. My counter-argument is that on a multilane un-divided carriageway, each lane-step to the right brings you closer to a lane that is uninterrupted enough to reach max capacity. At which step (Lane 4, 5 or 6?) is hard to say!

The other factor, at least in my experience, is that once you get above three lanes, a lot of drivers haven't the faintest idea which lane they should be in. Which, outside of peaks where every lane is full, serves to reduce the real-life capacity of the road. I drove along the M4 into London recently and for large distances the inner 2 lanes were virtually empty. I find that is common on 4 lane sections. That's an education problem, but no one wants to tackle it

It is simple enough - keep to Lane 1 unless overtaking - the idea of a "slow lane" and "fast lane" aren't particularly helpful.

Unfortunately, with the (perhaps overplayed) safety concerns about ALR, some motorists have opted to stay out of Lane 1 on such roads entirely, fearing they may come across a stopped vehicle.

[Bryn666]

Exactly. That weaving influence is the argument for C/D. I think uninterrupted express lanes do work at, or much nearer to, theoretical maximum capacity, so the C/D argument certainly has a case. My counter-argument is that on a multilane un-divided carriageway, each lane-step to the right brings you closer to a lane that is uninterrupted enough to reach max capacity. At which step (Lane 4, 5 or 6?) is hard to say!

The other factor, at least in my experience, is that once you get above three lanes, a lot of drivers haven't the faintest idea which lane they should be in. Which, outside of peaks where every lane is full, serves to reduce the real-life capacity of the road. I drove along the M4 into London recently and for large distances the inner 2 lanes were virtually empty. I find that is common on 4 lane sections. That's an education problem, but no one wants to tackle it

It is simple enough - keep to Lane 1 unless overtaking - the idea of a "slow lane" and "fast lane" aren't particularly helpful.

Unfortunately, with the (perhaps overplayed) safety concerns about ALR, some motorists have opted to stay out of Lane 1 on such roads entirely, fearing they may come across a stopped vehicle.

Even before smart motorways there were a depressingly high number of blissfully unaware numpties that would sit in lane 2 doing about 60 mph whilst traffic whizzed around them on both sides.

Driving is a serious skill, and it's more than apparent a vast number aren't up to it. Get these useless dangers to everyone else off the roads and the arguments about C/D lanes vs D5/6 become largely academic.

[Peter Freeman]

The other factor, at least in my experience, is that once you get above three lanes, a lot of drivers haven't the faintest idea which lane they should be in. Which, outside of peaks where every lane is full, serves to reduce the real-life capacity of the road. I drove along the M4 into London recently and for large distances the inner 2 lanes were virtually empty. I find that is common on 4 lane sections. That's an education problem, but no one wants to tackle it.

Poor driving behaviour - annoying, but is the issue overstated? Was it really so bad? Did it actually cause a problem for you (ie. affect your journey-time)? Was the general flow speed much below the limit? When I'm in the UK, it seems to me that most motorway right lane occupiers are over 70mph.

That frustrating behaviour is not reducing 'the real life capacity of the road'. Extra capacity is there after widening from 3 to 4 or 5 lanes, but it's not being used efficiently off-peak. It does get used efficiently on-peak, and that period is the reason for lanes being added.

'Right-laning' happens in Australia too, despite occasional signs advising "keep left except when overtaking". It's not specifically illegal, but I think the police could charge you under some convenient statute if they wished. Of course, they don't. However, we legally overtake on the left, if necessary (though the police could charge you for that too, if done unsafely). I'm not suggesting the UK should change. You just need to get used to the left lanes being used inefficiently off-peak. Consider them to be on hot standby, and conveniently available for joiners and leavers!

[ManomayLR]

Even before smart motorways there were a depressingly high number of blissfully unaware numpties that would sit in lane 2 doing about 60 mph whilst traffic whizzed around them on both sides.

Driving is a serious skill, and it's more than apparent a vast number aren't up to it. Get these useless dangers to everyone else off the roads and the arguments about C/D lanes vs D5/6 become largely academic.

Just the other day we were driving towards the M11 on the M25 ALR.

Broken down vehicle and it had been dealt with in a textbook perfect way.

Everyone slowed down, the Red X over lane 1, people moved past in an orderly manner and all four lanes opened straight afterwards.

Except for this one white van which pulled into the red X lane right in front of the gantry. Which had a camera. Obliviously drove on in the closed lane and, of course, got flashed.