Zipper Merge News

[kalvado]

I thought most drivers know 2 second rule - that means lane throughput of 1800 VPH.

Where did you come up with 1800? I pulled 600 VPH out of my ass, just as an example.

Oh, out of your ass... You need to stop giving us your shit...

WTF are you even talking about?????

Did you ever think about role of "ass" in humans? One of the things is that output of gastrointestinal tract is located there.
Food you consume travels through gastrointestinal tract, substances your body need are absorbed -  and leftovers are ejected in a form of feces, commonly known as "shit". It comes from "anus", locates between buttocks  - area commonly known as "ass".
So looks like there is only one type of substance that jakeroot could  pull out of his ass..

I was reacting to your commentary having had nothing to do with the topic at hand and turning into pointless rude ranting. The so-called "two-second rule" had nothing to do with anything. Time for you to learn some manners.

Well, if you think about 2 seconds as being random number - maybe.
If you think about human drivers as being somewhat sane and modestly responsible, you may realize most people are not willing to go well below 2 seconds.
There are numerous studies on the subject, and they all produce similar dependences:



maximum flow rate on a highway is about 1800-2000, maybe 2100 vehicles per hour. It slightly depends on road conditions - like speed limit (speed of free flow, to be more precise), number of lanes etc. So  2 second rule looks like an extension of  typical driver behavior.
Try to push things beyond that - and traffic falls into congested flow.

I thought anyone having some modest interest in road and traffic related issues should have seen those graphs more than once.

[jeffandnicole]

A lot of it comes down to the area that one lives in.  Here in NJ, late merging is almost necessary because of the traffic we have.  The other day, Rt. 55 approaching Rt. 42 was jammed for about 4 miles.  Do you have 2, 4 mile lanes of traffic, or one long 8 mile jam, including traversing thru 3 interchanges?  Around here, using all available lanes is not only common, but necessary.  In more rural areas where people aren't used to traffic jams, they probably want to get over a bit earlier.  And guaranteed, the ones using the open lane to the end are those used to more congested conditions!

Quick estimate gives 40-60 minutes worth of traffic in such backup.. I would say this qualifies as a parking lot!
And I would think at that point some active control measures - such as setting up detours - should start.. or there is no detour space as well?

Detours would simply jams up other roads, most of which are single lane roads that'll just jam up all the local traffic needlessly.  It was on the weekend, so it would take up resources and require overtime expenses.  In this particular case, it was a combination of shore traffic and rainy weather, and the jammed routes are relieving traffic from other heavily travelled highways already.  Not much that can be done about it.  People may grumble about the traffic, but it's just another normal day in paradise!

[1995hoo]

I think kalvado needs to drive on the roads on which I drive. I won't say people are engaged in NASCAR-style drafting. But I'd say people following the so-called "two-second rule" are the exception, not the rule.

[kalvado]

I think kalvado needs to drive on the roads on which I drive. I won't say people are engaged in NASCAR-style drafting. But I'd say people following the so-called "two-second rule" are the exception, not the rule.

Give FHWA a call, they may learn something new. So far, they consider 1850 VPH as LOS F...

[paulthemapguy]

maximum flow rate on a highway is about 1800-2000, maybe 2100 vehicles per hour. It slightly depends on road conditions - like speed limit (speed of free flow, to be more precise), number of lanes etc.

slightly

But yeah...the lack of ability to see eye-to-eye may largely be dependent on a difference in driving environments...with rural roads that never approach max capacity, merging is an entirely different animal.  Roads with tighter traffic need different considerations.  But another thing I've been seeing is excessive speculation--making 100 assumptions leading to a conclusion, which is presumed to be way more solid than it actually is.  (40 to 60 minute backup?  Where do you get that?)

My main take on zipper merges-Zipper merges are great, but not enforceable.  A zipper merge is a product of drivers choosing to take turns--it is not a necessary result of building infrastructure that fosters them.  Preferences toward a zipper merge vary regionally, because they exist due to driver culture.  At a particular merge zone, you can put up signs and stripe lanes in a way that makes it ambiguous which lane ends and which continues, but that's not going to necessarily steer people's behavior toward taking turns like we learned in kindergarten.  I've only really seen signage and traffic control spelling out zipper merges in Maine, where there's never really any tight traffic.  I think agencies in more urbanized areas realize the futility of signs and striping instructing motorists to enact a zipper merge.

[jeffandnicole]

I think agencies in more urbanized areas realize the futility of signs and striping instructing motorists to enact a zipper merge.

In more urbanized areas, it happens naturally.  There's no need for signage. 

[kalvado]

maximum flow rate on a highway is about 1800-2000, maybe 2100 vehicles per hour. It slightly depends on road conditions - like speed limit (speed of free flow, to be more precise), number of lanes etc.

slightly

But yeah...the lack of ability to see eye-to-eye may largely be dependent on a difference in driving environments...with rural roads that never approach max capacity, merging is an entirely different animal.  Roads with tighter traffic need different considerations.  But another thing I've been seeing is excessive speculation--making 100 assumptions leading to a conclusion, which is presumed to be way more solid than it actually is.  (40 to 60 minute backup?  Where do you get that?)

My main take on zipper merges-Zipper merges are great, but not enforceable.  A zipper merge is a product of drivers choosing to take turns--it is not a necessary result of building infrastructure that fosters them.  Preferences toward a zipper merge vary regionally, because they exist due to driver culture.  At a particular merge zone, you can put up signs and stripe lanes in a way that makes it ambiguous which lane ends and which continues, but that's not going to necessarily steer people's behavior toward taking turns like we learned in kindergarten.  I've only really seen signage and traffic control spelling out zipper merges in Maine, where there's never really any tight traffic.  I think agencies in more urbanized areas realize the futility of signs and striping instructing motorists to enact a zipper merge.

Well, if you noticed - jeffandnicole was describing a specific situation observed at specific time: 4 miles of backup on a 2-lane road. 8 lane-miles worth of cars.  A middle school student would be able to estimate how many cars are there (google average car length, assume certain gaps between cars),  and well - throughput of remaining lane on the order of 1.2-1.5 kVPH would give an estimate of waiting time.
So which exact assumption you think is random? They are not accurate, but they give me an idea of what was going on. And - which I find more useful - they give me some idea of how traffic irregularities are handled and can be handled in NJ.

Oh, and 1.8-2.1 kVPH is for a highway stretch, no merges.

Any more questions?

[vdeane]

Throughput of the road is limited by the spot of  least throughput.
did you ever hear words "weakest link"? This is exactly the case.

No shit. That's our point. The consensus by the morons in this thread seems to be that if you merge the moment you see a gap, there won't be a backup at the merge point. The only way to guarantee this not to be an issue, by their standards at least, is to completely close off the lane that ends, such that there simply is no opportunity to be in the lane that ends, hence, no merge. The problem is, that's fucking stupid: more lanes equal higher volume, period. A single lane of travel does not have an infinite vehicle-per-hour throughput. Cars still have to leave a reasonable gap, and the road still has a speed limit.

For example, If a single lane has a VPH throughput of 600 cars, three lanes have a throughput of 1800 cars. Now, this should be obvious, but it's a concept that seems to be ignored by those who merge early. Merging early does not increase the maximum efficiency of a lane -- that is determined by the speed limit, and the following distance of the cars in said lane.

Now, does a merge lower the maximum theoretical throughput of a lane? Of course. But sometimes, lanes have to end. And sometimes, you need multiple lanes to handle demand. It's a god damn balancing act, and there's no perfect solution here.

It's a judgement call, balancing finding the gap with not using the not-ending lane unnecessarily.  It takes awareness of traffic patterns and the cars on the road, both ahead and behind.  It requires one to find a gap large enough to move over and match the speed of traffic in the other lane such that the traffic already there is not affected.  A person unable to do this shouldn't be allowed to drive.

I tend to find that approaching the merge is where the big slowdowns are, not when the lane has ended.  Once the lane has ended, traffic generally gets up to speed - not as fast as freeflow, but generally a respectable amount.  Approaching the merge can be a parking lot, because invariably someone uses "late merging" not to take turns but instead to pass a bunch of people and force everyone else to stop (often literally) and let them in.

[jeffandnicole]

maximum flow rate on a highway is about 1800-2000, maybe 2100 vehicles per hour. It slightly depends on road conditions - like speed limit (speed of free flow, to be more precise), number of lanes etc.

slightly

But yeah...the lack of ability to see eye-to-eye may largely be dependent on a difference in driving environments...with rural roads that never approach max capacity, merging is an entirely different animal.  Roads with tighter traffic need different considerations.  But another thing I've been seeing is excessive speculation--making 100 assumptions leading to a conclusion, which is presumed to be way more solid than it actually is.  (40 to 60 minute backup?  Where do you get that?)

My main take on zipper merges-Zipper merges are great, but not enforceable.  A zipper merge is a product of drivers choosing to take turns--it is not a necessary result of building infrastructure that fosters them.  Preferences toward a zipper merge vary regionally, because they exist due to driver culture.  At a particular merge zone, you can put up signs and stripe lanes in a way that makes it ambiguous which lane ends and which continues, but that's not going to necessarily steer people's behavior toward taking turns like we learned in kindergarten.  I've only really seen signage and traffic control spelling out zipper merges in Maine, where there's never really any tight traffic.  I think agencies in more urbanized areas realize the futility of signs and striping instructing motorists to enact a zipper merge.

Well, if you noticed - jeffandnicole was describing a specific situation observed at specific time: 4 miles of backup on a 2-lane road. 8 lane-miles worth of cars.  A middle school student would be able to estimate how many cars are there (google average car length, assume certain gaps between cars),  and well - throughput of remaining lane on the order of 1.2-1.5 kVPH would give an estimate of waiting time.
So which exact assumption you think is random? They are not accurate, but they give me an idea of what was going on. And - which I find more useful - they give me some idea of how traffic irregularities are handled and can be handled in NJ.

Oh, and 1.8-2.1 kVPH is for a highway stretch, no merges.

Any more questions?

I didn't sit in the jam, but I drove over part of it.  For a 4 mile backup in this location with the conditions present, it's probably pretty accurate.

[jakeroot]

Approaching the merge can be a parking lot, because invariably someone uses "late merging" not to take turns but instead to pass a bunch of people and force everyone else to stop (often literally) and let them in.

Sigh.

People have to merge somewhere, Valerie. They can't all merge early. The earlier people merge, the longer the backup gets.

[kalvado]

Approaching the merge can be a parking lot, because invariably someone uses "late merging" not to take turns but instead to pass a bunch of people and force everyone else to stop (often literally) and let them in.

Sigh.

People have to merge somewhere, Valerie. They can't all merge early. The earlier people merge, the longer the backup gets.

Longer in terms of what? And when and why that matters? So far, the only argument I can buy so far is extension of backup into exit/interchange. Even then, backup must be long-term parking lot to actually be an issue - and likely will extend into critical areas anyway after a while.

[jakeroot]

Approaching the merge can be a parking lot, because invariably someone uses "late merging" not to take turns but instead to pass a bunch of people and force everyone else to stop (often literally) and let them in.

Sigh.

People have to merge somewhere, Valerie. They can't all merge early. The earlier people merge, the longer the backup gets.

Longer in terms of what? And when and why that matters? So far, the only argument I can buy so far is extension of backup into exit/interchange. Even then, backup must be long-term parking lot to actually be an issue - and likely will extend into critical areas anyway after a while.

Longer in terms of "more cars in one lane equals longer backup", assuming a reasonable following distance.

[kalvado]

Approaching the merge can be a parking lot, because invariably someone uses "late merging" not to take turns but instead to pass a bunch of people and force everyone else to stop (often literally) and let them in.

Sigh.

People have to merge somewhere, Valerie. They can't all merge early. The earlier people merge, the longer the backup gets.

Longer in terms of what? And when and why that matters? So far, the only argument I can buy so far is extension of backup into exit/interchange. Even then, backup must be long-term parking lot to actually be an issue - and likely will extend into critical areas anyway after a while.

Longer in terms of "more cars in one lane equals longer backup", assuming a reasonable following distance.

So what?
Is it good? bad? doesn't really matter?  And why?

[paulthemapguy]

Oh I get it...let's try explaining it this way.
Say there's a merge, with a continuing lane and a terminating lane.  If people in the terminating lane merge earlier, the continuing lane ends up with a disproportionately large amount of traffic volume, making them subject to a worse traffic jam than if they all waited until the merge point and merged evenly.  Merging early results in a greater delay for those vehicles who end up in the continuing lane prior to the merge point, with these vehicles now comprising a greater proportion of traffic.  Greater proportion of traffic means greater delays due to jamming.  And I would assume that greater delays are bad, right?

Any more questions?

[Scott5114]

Around OKC, there's plenty of afternoon rush along the 44 and the 35

* Scott5114 twitches

[jeffandnicole]

A lot of it comes down to the area that one lives in.  Here in NJ, late merging is almost necessary because of the traffic we have.  The other day, Rt. 55 approaching Rt. 42 was jammed for about 4 miles.  Do you have 2, 4 mile lanes of traffic, or one long 8 mile jam, including traversing thru 3 interchanges?  Around here, using all available lanes is not only common, but necessary.  In more rural areas where people aren't used to traffic jams, they probably want to get over a bit earlier.  And guaranteed, the ones using the open lane to the end are those used to more congested conditions!

Quick estimate gives 40-60 minutes worth of traffic in such backup.. I would say this qualifies as a parking lot!
And I would think at that point some active control measures - such as setting up detours - should start.. or there is no detour space as well?

After thinking about it last night, this statement shows the clear difference between those in rural areas where congestion is having to sit thru 2 traffic light cycles...and suburban/urban areas.

On my daily commute home, I will sit thru a 10 mile backup...on a GOOD day.  14 mile backups to the previous interchange aren't terribly uncommon.  The actual delay time isn't all that bad, maybe another 20 minutes thru this stretch because of how rush hour congestion works.  But it just goes to show that if a 4 mile backup makes one think an official detour is needed, I don't know what they would do in a 14 mile backup!

Occasionally there'll be small sections where you can get back up to the speed limit, and on unusual days you may even hit well above the speed limit.  But anymore, we just hope to slow down to a point where we're only losing 15 - 20 minutes due to congestion.

[kalvado]

Oh I get it...let's try explaining it this way.
Say there's a merge, with a continuing lane and a terminating lane.  If people in the terminating lane merge earlier, the continuing lane ends up with a disproportionately large amount of traffic volume, making them subject to a worse traffic jam than if they all waited until the merge point and merged evenly.  Merging early results in a greater delay for those vehicles who end up in the continuing lane prior to the merge point, with these vehicles now comprising a greater proportion of traffic.  Greater proportion of traffic means greater delays due to jamming.  And I would assume that greater delays are bad, right?

Any more questions?

Great! But wrong.
No, merging point does not affect wait time. Wait time depends on rate of arrival to a jam and rate of entrance into the problematic spot (assuming relatively free flow within remaining lane). You cannot change rate of arrival, and rate if entrance into the narrows is about same, likely higher for early merge. (Last statement is not very obvious, let's set it aside for now)
In simple words, if 60 cars arrive to the jam each minute but only 20 can proceed, your wait time will be increasing by 2 minutes for every minute of jam growth, no matter how they come to a stop.

[kalvado]

After thinking about it last night, this statement shows the clear difference between those in rural areas where congestion is having to sit thru 2 traffic light cycles...and suburban/urban areas.

On my daily commute home, I will sit thru a 10 mile backup...on a GOOD day.  14 mile backups to the previous interchange aren't terribly uncommon.  The actual delay time isn't all that bad, maybe another 20 minutes thru this stretch because of how rush hour congestion works.  But it just goes to show that if a 4 mile backup makes one think an official detour is needed, I don't know what they would do in a 14 mile backup!

Occasionally there'll be small sections where you can get back up to the speed limit, and on unusual days you may even hit well above the speed limit.  But anymore, we just hope to slow down to a point where we're only losing 15 - 20 minutes due to congestion.

Great point, actually.
But, if you will, there will be more about it in terms of growth of a jam.
I, maybe incorrectly, assume more or less steady flow of traffic. And 40 min jam is not just 40 min, it is an increase of, let's say, 20 min per hour. You seem to think about relatively short commute brust with some peak waiting time.

Beach traffic, from my perspective, should have longer period of high demand, and waiting time may end up growing until traffic starts diverting -e.g. when entrances become blocked.
But I am definitely in a group who believe that 2 light cycle wait is a beginning of a jam.

[jeffandnicole]

After thinking about it last night, this statement shows the clear difference between those in rural areas where congestion is having to sit thru 2 traffic light cycles...and suburban/urban areas.

On my daily commute home, I will sit thru a 10 mile backup...on a GOOD day.  14 mile backups to the previous interchange aren't terribly uncommon.  The actual delay time isn't all that bad, maybe another 20 minutes thru this stretch because of how rush hour congestion works.  But it just goes to show that if a 4 mile backup makes one think an official detour is needed, I don't know what they would do in a 14 mile backup!

Occasionally there'll be small sections where you can get back up to the speed limit, and on unusual days you may even hit well above the speed limit.  But anymore, we just hope to slow down to a point where we're only losing 15 - 20 minutes due to congestion.

Great point, actually.
But, if you will, there will be more about it in terms of growth of a jam.
I, maybe incorrectly, assume more or less steady flow of traffic. And 40 min jam is not just 40 min, it is an increase of, let's say, 20 min per hour. You seem to think about relatively short commute brust with some peak waiting time.

Beach traffic, from my perspective, should have longer period of high demand, and waiting time may end up growing until traffic starts diverting -e.g. when entrances become blocked.

I've never seen entrances blocked to a highway that's otherwise open but jammed.  Even with a highway that's shut down due to a major accident it's extremely rare for entrances to be blocked.  About the only time an entrance is blocked is when there's a motorcade or some other security-type event going on.

[Brandon]

Approaching the merge can be a parking lot, because invariably someone uses "late merging" not to take turns but instead to pass a bunch of people and force everyone else to stop (often literally) and let them in.

Sigh.

People have to merge somewhere, Valerie. They can't all merge early. The earlier people merge, the longer the backup gets.

Longer in terms of what? And when and why that matters? So far, the only argument I can buy so far is extension of backup into exit/interchange. Even then, backup must be long-term parking lot to actually be an issue - and likely will extend into critical areas anyway after a while.

Longer in terms of "more cars in one lane equals longer backup", assuming a reasonable following distance.

Why should there be a backup?  If people merge at any given time at speed, and into an appropriate gap (as long as others are following at a reasonable distance instead of tailgating), then there need be no real backup at whatever point the merge occurs.  The problem comes in, in that people tailgate, actively block others from merging, while others bull their way in.  This is the crappola that happens in Chicago on a daily basis.  It's not road design, it's asinine driver behavior.

[jeffandnicole]

Approaching the merge can be a parking lot, because invariably someone uses "late merging" not to take turns but instead to pass a bunch of people and force everyone else to stop (often literally) and let them in.

Sigh.

People have to merge somewhere, Valerie. They can't all merge early. The earlier people merge, the longer the backup gets.

Longer in terms of what? And when and why that matters? So far, the only argument I can buy so far is extension of backup into exit/interchange. Even then, backup must be long-term parking lot to actually be an issue - and likely will extend into critical areas anyway after a while.

Longer in terms of "more cars in one lane equals longer backup", assuming a reasonable following distance.

Why should there be a backup?  If people merge at any given time at speed, and into an appropriate gap (as long as others are following at a reasonable distance instead of tailgating), then there need be no real backup at whatever point the merge occurs.  The problem comes in, in that people tailgate, actively block others from merging, while others bull their way in.  This is the crappola that happens in Chicago on a daily basis.  It's not road design, it's asinine driver behavior.

The backup doesn't occur because people are merging...the backup is occurring because of a blockage further down the road.  In a recent case I many have mentioned, traffic was stopped in the construction zone due to work going on for a few minutes.  Traffic at the merge point can't go the speed limit if traffic is stopped.

Even if traffic was moving at speed, in order for everyone to have an appropriate distance prior to the merge, they would need 4 seconds between vehicles.  As they zipper in, then it would be 2 seconds between vehicles.

[kalvado]

After thinking about it last night, this statement shows the clear difference between those in rural areas where congestion is having to sit thru 2 traffic light cycles...and suburban/urban areas.

On my daily commute home, I will sit thru a 10 mile backup...on a GOOD day.  14 mile backups to the previous interchange aren't terribly uncommon.  The actual delay time isn't all that bad, maybe another 20 minutes thru this stretch because of how rush hour congestion works.  But it just goes to show that if a 4 mile backup makes one think an official detour is needed, I don't know what they would do in a 14 mile backup!

Occasionally there'll be small sections where you can get back up to the speed limit, and on unusual days you may even hit well above the speed limit.  But anymore, we just hope to slow down to a point where we're only losing 15 - 20 minutes due to congestion.

Great point, actually.
But, if you will, there will be more about it in terms of growth of a jam.
I, maybe incorrectly, assume more or less steady flow of traffic. And 40 min jam is not just 40 min, it is an increase of, let's say, 20 min per hour. You seem to think about relatively short commute brust with some peak waiting time.

Beach traffic, from my perspective, should have longer period of high demand, and waiting time may end up growing until traffic starts diverting -e.g. when entrances become blocked.

I've never seen entrances blocked to a highway that's otherwise open but jammed.  Even with a highway that's shut down due to a major accident it's extremely rare for entrances to be blocked.  About the only time an entrance is blocked is when there's a motorcade or some other security-type event going on.

By "blocked" I mean "blocked by traffic", not "blocked by police officers"... I hope that makes more sense?

[Brandon]

The backup doesn't occur because people are merging...the backup is occurring because of a blockage further down the road.  In a recent case I many have mentioned, traffic was stopped in the construction zone due to work going on for a few minutes.  Traffic at the merge point can't go the speed limit if traffic is stopped.

Bullshit.  As soon as we're past the merge point around here, traffic speeds up.  The merge point IS the blockage point around Chicago.

[DaBigE]

Bullshit.  As soon as we're past the merge point around here, traffic speeds up.  The merge point IS the blockage point around Chicago.

Seconded. Anytime I've gone thru the Zoo Interchange construction zone zipper, traffic has sped up once we were out of the zipper.

[jeffandnicole]

After thinking about it last night, this statement shows the clear difference between those in rural areas where congestion is having to sit thru 2 traffic light cycles...and suburban/urban areas.

On my daily commute home, I will sit thru a 10 mile backup...on a GOOD day.  14 mile backups to the previous interchange aren't terribly uncommon.  The actual delay time isn't all that bad, maybe another 20 minutes thru this stretch because of how rush hour congestion works.  But it just goes to show that if a 4 mile backup makes one think an official detour is needed, I don't know what they would do in a 14 mile backup!

Occasionally there'll be small sections where you can get back up to the speed limit, and on unusual days you may even hit well above the speed limit.  But anymore, we just hope to slow down to a point where we're only losing 15 - 20 minutes due to congestion.

Great point, actually.
But, if you will, there will be more about it in terms of growth of a jam.
I, maybe incorrectly, assume more or less steady flow of traffic. And 40 min jam is not just 40 min, it is an increase of, let's say, 20 min per hour. You seem to think about relatively short commute brust with some peak waiting time.

Beach traffic, from my perspective, should have longer period of high demand, and waiting time may end up growing until traffic starts diverting -e.g. when entrances become blocked.

I've never seen entrances blocked to a highway that's otherwise open but jammed.  Even with a highway that's shut down due to a major accident it's extremely rare for entrances to be blocked.  About the only time an entrance is blocked is when there's a motorcade or some other security-type event going on.

By "blocked" I mean "blocked by traffic", not "blocked by police officers"... I hope that makes more sense?

Yeah...and that does happen a bit more often.  Although even then it depends.  If it's a movement like (for example) NJ 73 South to 295 South and 295 South to 73 South, that one is jammed so often that most people just naturally assume it's jammed due to regular traffic.  In other cases, a jam on the ramp would be more unusual, and you'll be seeing people back up off the ramp.  In extreme cases which doesn't happen often at all, you'll see people going the wrong way, off an on-ramp (and without any cop actually allowing them to do so!).