Traffic theory: merge in turn or queue?

Factual Questions
1

First off, I’m not at all sure where this thread belongs; it might be a debate, it might just be opinion, or there might be a factual answer if studies have been done.

On the way to work this morning, I was delayed by a traffic queue backing up from some road works where two lanes were reduced to one.

In the UK, this is typically announced by lane closed ahead signs (NB: this image shows a sign warning the closure of two lanes out of four), repeated at 800, 600, 400, 200 and 100 yards prior to the point of closure.

Typically, the majority of drivers will try to get into the lane which will remain open and they will do so as soon as they see the 800 yard sign - this results in stationary queues and 800 yards of empty road in the lane that will close. A few bold drivers do not try to queue, but drive right up the empty lane and attempt to merge at the point where the lane closes - this behaviour is almost universally resented by the other (queueing) drivers.
Some drivers (particularly those of larger vehicles) attempt to thwart the non-queuers by straddling the central line, others simply refuse to let them merge, resulting in heated tempers all round.

The question:
Would it be better if the warning signs said 'lanes reduced in (x) yards: merge STRICTLY IN TURN". - allowing both lanes to be utilised for queueing? - reducing the tailback length by half and thus reducing the probability that this will back up far enough to interfere with another junction.
Or is it the case that there would be more minor accidents in the case of merge-in-turn?

2

The answer to this question actually depends as much on human nature as on traffic or queuing theory.

There are plenty of models which demonstrate the flow of traffic along a road dependent on factors such as maximum speed and number of lanes. It’s the use of this kind of model which dictates that it’s good sense to warn people to slow from a normal maximum speed to a lesser maximum speed when there’s an upcoming obstruction or lane restriction.

The problem is that with human nature being what it is, people will ignore the warnings to slow down until they actually see the obstruction and a queue forms which may actually have been avoided if everyone acted for the common good and slowed down.

In the ideal world, people would slow down immediately upon instruction and merge gradually to enable the slower, narrower stream of traffic to flow smoothly through the restriction. What actually happens is that people race to the obstruction and therefore cause a jam.

Forcing people to merge in turn would actually exacerbate the problem (in theory). By making people merge closer to the restriction, you’re forcing them to slow down and filter in a smaller space - which obviously holds greater risk of accident. This abrupt slowing of traffic would also increase the likelihood of the queue forming in the first place!

In practice, of course, the queues generally do form for exactly this reason - people refusing to slow down and attempting to merge at the last minute, thereby forcing others to slow to allow them in. The question then is, once the queue has formed, is it better to have two lines of traffic for as great a time as possible?

The simple answer is that it should not make any difference to the rate at which traffic progresses once the queues have formed. Obviously, it would reduce the length of the queue and that in itself has advantages, as you noted.

The down side is the question of when to switch from the optimistic view that ignores human nature and hopes for the best (also avoiding the risk of higher speed collisions) to the pessimistic view which tries to minimise the length of the queue.

For the record, queuing theory is an established area of mathematical study and traffic theory also has quite a body of study. Traffic theory has derived rules for all manner of things such as methods to calculate “inter-green time” for traffic lights (the time between one set of lights being green and the next - minimised to allow traffic flow without risking accidents).

My days of studying these things were some ten years ago and I may have been asleep at the time. I’m sure someone will put any mistakes right.

Stu Hern

3

I just want to add that German traffic law was changed in this matter about three years ago. Before you were supposed to merge about IIRC 200 metres before the end of the closing lane. Now you are supposed to stay on the closing lane as long as possible.
I can tell you there was a lot of heated temper when people acted according to the new law and many people didn’t know there was a change. Now things have calmed down and it works rather smoothly. I like it better than before when you had to merge very early and than had some BMW or Mercesdes zoom past you on the empty lane trying to avoid another 500 metres of slow driving.

4

Saw a sign (actually multiple) signs in PA stating exactly that, well the words were different but the meanign is the same. IIRC “Stay in lane till merge point, then alternate”

5

This same discussion has been a raging debate in the D.C. area carried out in the local newspaper. Traffic here is so bad we have a regular column in the Post about traffic 2-3 times per week. The consensus seems to be, theory aside, that the alternate merge (also called the “zipper merge”) at the point of lane loss is the smoothest approach. But it requires the driving culture to support it–everybody has to play. Everybody who has written with experience with this type of merge behavior says it produces the best results.

However, this is not a scientific study, it’s anecdotal.

6

Merging does work! The basic problem lies with the English mentality. An example:

I’ve lived for the last 11 years in Hungary , where, although the standard of driving leaves a lot to be desired, and the law of the jungle often applies, it’s generally accepted that if everyone lets one car in front of them, the two lines of cars merge in a zipper-like fashion. And it actually works most of the time.

Two years ago I drove back to England for a visit, and about 10 minutes after driving off the ferry I came to a section of roadworks on the motorway. Naively - having been out of the country too long, I guess - I allowed one car to merge in front of me, and…

All hell broke loose!

The cars behind this guy drove up to each other’s bumpers to form a solid impenatrable line and cut in front of me. Okay, I says: if that’s the way you want to play it! And, with space running out before I’d have had to brake to allow the pushy bastards in front of me, I edged forward to stem the flow.

But the stupid, arrogent prick in the car I’d chosen actually took offense at this and refused to get behind me even though we were now neck and neck in a single lane with traffic cones brushing our flanks. In addition, the besuited wanker was honking his horn and flicking v-signs through the window while - as far as I could tell from my attempts to lipread through his wondow - telling his wife and screaming kids to “shut the f*ck up.”

His car being marginally faster than mine, I eventually let him in front and had to make do with the lesser satisfaction of driving behind him with my lights on full beam for the remaining 20 miles or so of contra-flow.

Whatever, the upshot is that drivers in England are incapable of grasping the concept of considerate driving, which is a shame, because they have some of the best roads in Europe, possibly the world.

7

And no, I wasn’t in my trabi. I sold that about six months ago.

Soory, not on the subject I know, just a pretext to see if I could figure out how to insert links in posting.

8

As an Operations problem, it doesn’t really matter how the queue forms, because the eventual one-lane passageway becomes the bottleneck. Bottlenecks pace the system, so the speed before it does not matter.

9

Let me introduce you to the word ‘some’.

10

In that case I should inform you of the existence of another word: ‘most’.

We are talking about the home of white van man and road rage after all.

Also, just in the instance I described (and in a former life I have been a taxi driver, motorbike courier and van driver in England, so this was not exactly my first experience of bad driving there), not one of the vast number of cars in my immediate vicinity displayed the slightest intention of “merging,” or even giving a flying f*ck about anyone else, as they were all too busy trying to gain an extra 10 yards of road, presumably in the hope of arriving home a few nanoseconds early.

Having said that, mangetout, the fact that you have even thought about the matter enough to post a question about it probably makes you one of the notable exceptions.

11

I disagree with your usage of ‘most’; don’t get me wrong; there are plenty of bad, aggressive, stupid drivers, but I think its just a case of the bad minority making themselves noticed more.

it is possible that your perspective as a motorcyclist has coloured your perceptions too…

I’m pretty confident that most of the aggression in the traffic queue I mentioned is the result of frustration and confusion caused by unclear instructions - if the signs said ‘merge before 600 yards’ or if they said ‘stay in lane and merge at end’, I reckon people would be a lot more comfortable and generous about it - we actually like to be told what to do - we like the boundaries to have been defined and the ignorance and confusion that occurs when they aren’t can look like malice.

12

Frustration and agression yes, lack of clear instructions - I don’t think so. It probably has more to do with the frustration caused by overcrowding and constant traffic jams on Britain’s roads.

Also, maybe people are more laid back where I live and I’ve just got used to it, but I generally find the English to be increasingly agressive and frustrated anyway. Especially at chucking out time. Is it just me, or has the atmosphere really turned nastier over there?

13

Dunno; were you in London by any chance?

14

I usually split my time between London and Cambridge (where I grew up) when I visit.

Sorry, I’ve got to go away for a couple of days now, so I can only continue this discussion on Friday.

Drive carefully in the meantime!

15

It depends.
During times of heavy traffic where the road backs up regardless, then the German rule seems appropriate. However, when traffic is light and the traffic would continue to flow by merging early then that would be the better method.

16

During rush hour in the Seattle area there are several merge zones which cause traffic-backups (tailbacks), and it’s the merge-zone itself which causes the bottleneck. The single lane after the merge zone is not the problem. Once past the merge zone, the spaces between cars in the single lane are absolutely enormous.

If the incoming lanes pack together and only merge at the last minute, then merging resembles the traffic at a 4-way stop sign: one driver halts completely while the driver in the other lane takes his turn to move forward.

See the “merging lane traffic jams” section of my article:

TRAFFIC WAVES