M5 Traffic Jams Southbound past J20 (Clevedon)

[CrazyInWeston]

Having said that however, why then going northbound the same thing (J20 adds a lane and again becomes the new lane number 1) doesnt happen? (I say this as every time I drive it I'm actually more likely to be stuck in a jam going north after the lane drop up the hill than actually going up the hill, theres no bottle neck of lane 4 going to lane 3.) People travelled south, they must travel back home after all. I see more traffic jams northbound on the downhill section thats just before J19 than the supposedly increased capacity of 4 lanes when going past J20 north.

I would guess that significantly more traffic joins the M5 southbound at J19 than northbound at J20, so traffic can shuffle over more effectively going north

You would think that, but between 6 to 10am northbound J21 there is literally a steady stream of traffic entering the Motorway forcing drivers already on it to move from lane 1 to lane 2 as there are that many cars joining and during this time there doesnt seem to be an impact or slow down due to this steady stream of cars entering the Motorway northbound at J21.

[Leaferclove]

Sorry for dragging this old thread up, but I found WHBM's explanation of why there's the wide central reservation between J20 and J21 of the M5 quite interesting. Whether or not it's correct, or as RichardA35 says, in fact refers to a section in the Taunton area, this piece I can corroborate:

There was a special railway siding installed during construction, probably somewhere near where the two cross, which was used to deliver the considerable quantity of heavy fill that went into this section.

Not far to the west from where the Bristol to Taunton line runs under the M5 (just north of J21) is the location of the former Puxton and Worle station (where Station Road meets the line at the level crossing). I understand temporary sidings were built at this location while the M5 was being built for trains bringing the stone for construction. By that point, the station had been closed by Beeching. I'm sure using trains in this manner was probably a fairly common occurrence for road building back in the day, but I thought I'd chip in with that little anecdote nonetheless!

[SteveA30]

They were known as the Puxton fly-ash trains and ran approx 1970-72ish. Possibly from a South Wales location. I may have old rail magazines with the news item somewhere....

This would seem to be it, in the notes underneath
https://www.flickr.com/photos/blue-diesels/34633837220

[Greg07]

Saw this thread suddenly appear and can't resist putting a link to a video ( YouTube) I took of this M5 summer queue back in 2021 between junctions 19 and 20 on the split level section where I noticed the queue formed interesting ( or maybe not but .. ) successive "queue waves" where the queue could be seen to propagate along then clear along the motorway in clear wave movements.

https://youtu.be/71kzjXUdsH4

[DB617]

Saw this thread suddenly appear and can't resist putting a link to a video ( YouTube) I took of this M5 summer queue back in 2021 between junctions 19 and 20 on the split level section where I noticed the queue formed interesting ( or maybe not but .. ) successive "queue waves" where the queue could be seen to propagate along then clear along the motorway in clear wave movements.

https://youtu.be/71kzjXUdsH4

This is a great real life demonstration of why modelling traffic as fluid flow does not work - driver perception plays too much of a part in traffic speeds. You can't see the traffic slowing ahead quite often, and a certain number of drivers would probably not react appropriately (by slowing down so as not to close the gap ahead). I try to do it where I can, but on a multi-lane carriageway, inevitably this results in people joining your lane thinking it will get them somewhere.

Trying to make progress or even make your own life smoother on a motorway suffering from flow breakdown is as demoralising and painful an experience as a driver can have with an intact car.

[Steven]

Sorry for dragging this old thread up, but I found WHBM's explanation of why there's the wide central reservation between J20 and J21 of the M5 quite interesting. Whether or not it's correct, or as RichardA35 says, in fact refers to a section in the Taunton area, this piece I can corroborate:

Which is interesting and all, but it's not actually true...

It has the extra wide central reservation to allow the road to be easily widened to D4M at a later stage to account for the expected additional traffic flows between the South Bristol Spur at J20 and the Weston-super-Mare Spur at J21; neither of which came to fruition.

It's the same idea as the original extra-wide central reservation on the Preston Bypass.

[Norfolktolancashire]

Saw this thread suddenly appear and can't resist putting a link to a video ( YouTube) I took of this M5 summer queue back in 2021 between junctions 19 and 20 on the split level section where I noticed the queue formed interesting ( or maybe not but .. ) successive "queue waves" where the queue could be seen to propagate along then clear along the motorway in clear wave movements.

https://youtu.be/71kzjXUdsH4

This is a great real life demonstration of why modelling traffic as fluid flow does not work - driver perception plays too much of a part in traffic speeds. You can't see the traffic slowing ahead quite often, and a certain number of drivers would probably not react appropriately (by slowing down so as not to close the gap ahead). I try to do it where I can, but on a multi-lane carriageway, inevitably this results in people joining your lane thinking it will get them somewhere.

Trying to make progress or even make your own life smoother on a motorway suffering from flow breakdown is as demoralising and painful an experience as a driver can have with an intact car.

Agreed, trying to make progress along such a motorway is frustrating. Many drivers believe that by remaining in lane 3/4 results in less travelling time, incorrectly on a "congested" motorway such as this section of the M5 during peak times.
A bit of an eye opener for me was when I drove a rental van southbound on this stretch, remaining at around 60mph in lane 1. The whole journey from Lancashire to Cornwall was not much longer in time than trying to maintain 70mph in my own car, and a lot less fraught!

[jnty]

Saw this thread suddenly appear and can't resist putting a link to a video ( YouTube) I took of this M5 summer queue back in 2021 between junctions 19 and 20 on the split level section where I noticed the queue formed interesting ( or maybe not but .. ) successive "queue waves" where the queue could be seen to propagate along then clear along the motorway in clear wave movements.

https://youtu.be/71kzjXUdsH4

This is a great real life demonstration of why modelling traffic as fluid flow does not work - driver perception plays too much of a part in traffic speeds. You can't see the traffic slowing ahead quite often, and a certain number of drivers would probably not react appropriately (by slowing down so as not to close the gap ahead). I try to do it where I can, but on a multi-lane carriageway, inevitably this results in people joining your lane thinking it will get them somewhere.

Trying to make progress or even make your own life smoother on a motorway suffering from flow breakdown is as demoralising and painful an experience as a driver can have with an intact car.

Agreed, trying to make progress along such a motorway is frustrating. Many drivers believe that by remaining in lane 3/4 results in less travelling time, incorrectly on a "congested" motorway such as this section of the M5 during peak times.
A bit of an eye opener for me was when I drove a rental van southbound on this stretch, remaining at around 60mph in lane 1. The whole journey from Lancashire to Cornwall was not much longer in time than trying to maintain 70mph in my own car, and a lot less fraught!

It's amazing how you can sometimes find yourself in lane 1 able to make quite good progress without passing anyone in lane 2 on the inside...because everyone is crammed into lane 3/4.

[swissferry]

When the M80(N) is congested between junctions 4a and 7 using lane 1 is usually only marginally slower than lane 2. Lane 1 tends to flow at a more constant speed whereas lane 2 is more start-stop. I generally use lane 1 as its more relaxing and economical.

[Peter Freeman]

Saw this thread suddenly appear and can't resist putting a link to a video ( YouTube) I took of this M5 summer queue back in 2021 between junctions 19 and 20 on the split level section where I noticed the queue formed interesting ( or maybe not but .. ) successive "queue waves" where the queue could be seen to propagate along then clear along the motorway in clear wave movements.

https://youtu.be/71kzjXUdsH4

This is a great real life demonstration of why modelling traffic as fluid flow does not work - driver perception plays too much of a part in traffic speeds. You can't see the traffic slowing ahead quite often, and a certain number of drivers would probably not react appropriately (by slowing down so as not to close the gap ahead). I try to do it where I can, but on a multi-lane carriageway, inevitably this results in people joining your lane thinking it will get them somewhere.

Trying to make progress or even make your own life smoother on a motorway suffering from flow breakdown is as demoralising and painful an experience as a driver can have with an intact car.

This is a great video, and that M5 split-level location is one of my favourite spots on the UK motorway network. The video clearly shows what some of us realise about dense traffic behaviour, but which the average driver will never have thought about.

Gradients, and heavy input at an on-ramp, can be triggers for flow breakdown. However, it will occur spontaneously at extreme traffic densities anyway. No need for a trigger.

The backward-moving 'Q-wave' (great name!) is the important part. It forms because drivers who see a slow-down in front of them are reluctant to slow down early enough. In effect, they unintentionally 'rush up to join the queue', and often come to a complete halt. When the car in front starts moving, there's a delay until the follower moves, and the gap opens wider than necessary. Before long the car in front is receding from you, and you speed up - but too much! Soon you've caught up with him, just as he's slowing down again, and this time the speed differential is larger. 'Rinse and repeat'.

Take-aways are -

1. VSL can help here, because it inhibits the 'speeding up too much' component of the process. (Primarily though, it defers flow breakdown beginning in the first place).

2. Adaptive cruise control, and autonomously driving cars, would probably not suffer this quite so severely as humans (not that I'm an advocate of autonomous vehicles).

3. Even without dense traffic or flow breakdown or these Q-waves, traffic tends to moves along motorways in 'platoons'. This is not evident from the driving seat, but it is when observed from an overbridge. The platoons are not permanent - their members interchange, but they're usually present. This is why metering of motorway-to-motorway connectors, not only on-ramps, works. Immediately downstream of the signal, where merging occurs, the joining traffic is briefly in that rare consistent-flow state, not platooned.

(note: Many commentators do not believe ramp-metering to be effective. Even more commentators, understandably, doubt the logic in metering M2M connectors, but this natural platooning phenomenon is the logic).