Zipper Merge News

[Revive 755]

Based on the message boards, it appears DuPage County, Illinois is trying a zipper merge on Winfield Road for bridge work at Mack Road.

[1995hoo]

Yesterday we encountered this sign on I-70 passing Terre Haute, Indiana. Wasn't a true zipper merge because the left lane ultimately ended more than a mile further down the road, but the sign is clear enough, right? Not to the people on I-70. Almost everyone panicked and moved over a mile in advance of the merge point. On the plus side for us, that meant we sailed past a long line of cars because I merge at the end just like the sign said (only one guy from Oklahoma seemed to take umbrage, but that was right at the end anyway).

[theline]

^^ During our recent trip on SR 37 near Bloomington, IN (on the I-69 upgrade), I saw signs like this. I had never before seen signs instructing drivers to zipper merge. The usual practice in Indiana has been to merge soon after the first signs warning of the lane closure.

The signs on 37 seemed to have some effect, but as 1995hoo says most drivers merged near the first warning.

SR 37 also had signs closer to the lane closure saying "MERGE HERE" or words to that effect. Few cars made it that far before merging.

[jakeroot]

I do wish the MUTCD would scale back regulations that require X-amount of warning before lanes end or merge. If we didn't advertise lane endings so early, we might get more zippering.

[kalvado]

I do wish the MUTCD would scale back regulations that require X-amount of warning before lanes end or merge. If we didn't advertise lane endings so early, we might get more zippering.

And point of all that being?...

[jakeroot]

I do wish the MUTCD would scale back regulations that require X-amount of warning before lanes end or merge. If we didn't advertise lane endings so early, we might get more zippering.

And point of all that being?...

I'm not going through this again. You know we have a difference of opinion on merging.

[kalvado]

I do wish the MUTCD would scale back regulations that require X-amount of warning before lanes end or merge. If we didn't advertise lane endings so early, we might get more zippering.

And point of all that being?...

I'm not going through this again. You know we have a difference of opinion on merging.

I still wonder if "better use of road capacity" means something more than " blahblahblah in lalaland" to anyone talking about the subject..

[jakeroot]

I do wish the MUTCD would scale back regulations that require X-amount of warning before lanes end or merge. If we didn't advertise lane endings so early, we might get more zippering.

And point of all that being?...

I'm not going through this again. You know we have a difference of opinion on merging.

I still wonder if "better use of road capacity" means something more than " blahblahblah in lalaland" to anyone talking about the subject..

It probably means something in construction zones, which are often more congested due to lane-drops, lower limits and narrowed lanes. Anything to improve capacity, you know?

[kphoger]

not going through this again

I down with this.

[kalvado]

I do wish the MUTCD would scale back regulations that require X-amount of warning before lanes end or merge. If we didn't advertise lane endings so early, we might get more zippering.

And point of all that being?...

I'm not going through this again. You know we have a difference of opinion on merging.

I still wonder if "better use of road capacity" means something more than " blahblahblah in lalaland" to anyone talking about the subject..

It probably means something in construction zones, which are often more congested due to lane-drops, lower limits and narrowed lanes. Anything to improve capacity, you know?

Well, before going to matters as complicated as "construction zone",  and "congestion", defining "capacity" may be a good step forward.

[paulthemapguy]

Well, before going to matters as complicated as "construction zone",  and "congestion", defining "capacity" may be a good step forward.

Vehicles per hour DOES NOT EQUAL amount of space occupied by vehicles at any given time.
Your definition of capacity scales upward with the amount of space occupied by vehicles.  Jake's definition (and mine) relates to the number of vehicles which can pass through a point on the roadway during a unit of time.  These definitions are not directly related.  Roadway capacity is measured in vehicles per hour, not vehicles per square yard.

[kalvado]

Well, before going to matters as complicated as "construction zone",  and "congestion", defining "capacity" may be a good step forward.

Vehicles per hour DOES NOT EQUAL amount of space occupied by vehicles at any given time.
Your definition of capacity scales upward with the amount of space occupied by vehicles.  Jake's definition (and mine) relates to the number of vehicles which can pass through a point on the roadway during a unit of time.  These definitions are not directly related.  Roadway capacity is measured in vehicles per hour, not vehicles per square yard.

You'll be surprised how many people on this forum do disagree with VPH metrics.. 
But once you adopt that metric, entire zipper concept falls apart.

[UCFKnights]

Back when I was in high school, they built a new building and did not expand the parking lot (parking capacity was not a real issue, before the expansion the lot would NEVER come close to filling up). After this, they ran into an issue where at pickup time, it would back up onto the main road and cause all sorts of problems, once it backed up onto it, it'd back up into other intersections and create gridlock. To try to solve this, they converted the parking lanes leading to the pickup area to be one way and "two lanes", and put signs at the beginning of this they put a sign "USE BOTH LANES" and at the end of it they put a sign "MERGE HERE - ALTERNATE LANES". The parents mostly ignored the signs, so they assumed they didn't see them and made additional signs, and installed 2 more of the same signs at the beginning and end, except one on the right side of the road, one in the center between the two lanes, and on on the left side. The parents still ignored the signs... they then decided to station the school officer by the sign to help people alternate between the two lanes... as soon as people were 50 feet past the person, they'd try to move back to the right lane way before the end of it. They added "STAY IN YOUR LANE" signs after the USE BOTH LANE signs and they didn't help. They started laying out cones to keep people in the lanes until the end which seemed effective, but any time someone wasn't out to direct the traffic, almost everyone would inevitably end up in the right lane and it'd backup onto the main road. When I asked parents why they wouldn't use the left lane, they said it wasn't going to make anything any faster, the pickup area only moves so fast, and they needed to be in the right lane for pickup anyways.

The length of the queue wasn't really off by too much... if they got about 10-15 cars into the left lane and people to pull forward and not leave multiple car lengths, the entire queue would be within the school property and it'd prevent the backups onto other roads. After 2 or 3 years of this, they redid the parking lot to add an extra loop to the queue so the cars could remain in a single lane.

So whats the point of this story? Its literally demonstrating the purpose of the zipper merge. So back to the topic:

You'll be surprised how many people on this forum do disagree with VPH metrics.. 
But once you adopt that metric, entire zipper concept falls apart.

Thats because VPH is the WRONG metric of what the zipper concept is trying to solve. The VPH through the merge point is ultimately what the road is capable of AFTER THE MERGE IS OVER, regardless of where the the merge point is.... that is the bottleneck and the zipper merge isn't going to improve that.

The proper metric would be linear feet of congestion. Does the zipper merge effectively decrease the number of linear feet of congestion? Because if an early merge is causing a 5 mile backup and switching it to a zipper merge reduces the backup by 1 mile, that means that anyone exiting in that one mile that the zipper merge reduces the congestion will now no longer experience any congestion, have their travel time cut down, and reduce the congestion for everyone by getting less people in the congested areas.

If you insist on using VPH as your metric, you actually can use that as well.... you just need to make sure you use it at the right spot on the road. The VPH metric will only be increased by measuring before each exit prior to the merge point. The VPH will mostly be increased at the first exit after the congestion would otherwise begin, with smaller increases as you get closer to the zipper merge. Using VPH close to the actual merge point is a meaningless metric, thats not where the zipper merge is designed to increase it.

When people fail to use the road capacity, the length of congestion increases. That is the metric that you are supposed to be using

[kalvado]

Back when I was in high school, they built a new building and did not expand the parking lot (parking capacity was not a real issue, before the expansion the lot would NEVER come close to filling up). After this, they ran into an issue where at pickup time, it would back up onto the main road and cause all sorts of problems, once it backed up onto it, it'd back up into other intersections and create gridlock. To try to solve this, they converted the parking lanes leading to the pickup area to be one way and "two lanes", and put signs at the beginning of this they put a sign "USE BOTH LANES" and at the end of it they put a sign "MERGE HERE - ALTERNATE LANES". The parents mostly ignored the signs, so they assumed they didn't see them and made additional signs, and installed 2 more of the same signs at the beginning and end, except one on the right side of the road, one in the center between the two lanes, and on on the left side. The parents still ignored the signs... they then decided to station the school officer by the sign to help people alternate between the two lanes... as soon as people were 50 feet past the person, they'd try to move back to the right lane way before the end of it. They added "STAY IN YOUR LANE" signs after the USE BOTH LANE signs and they didn't help. They started laying out cones to keep people in the lanes until the end which seemed effective, but any time someone wasn't out to direct the traffic, almost everyone would inevitably end up in the right lane and it'd backup onto the main road. When I asked parents why they wouldn't use the left lane, they said it wasn't going to make anything any faster, the pickup area only moves so fast, and they needed to be in the right lane for pickup anyways.

The length of the queue wasn't really off by too much... if they got about 10-15 cars into the left lane and people to pull forward and not leave multiple car lengths, the entire queue would be within the school property and it'd prevent the backups onto other roads. After 2 or 3 years of this, they redid the parking lot to add an extra loop to the queue so the cars could remain in a single lane.

So whats the point of this story? Its literally demonstrating the purpose of the zipper merge. So back to the topic:

You'll be surprised how many people on this forum do disagree with VPH metrics.. 
But once you adopt that metric, entire zipper concept falls apart.

Thats because VPH is the WRONG metric of what the zipper concept is trying to solve. The VPH through the merge point is ultimately what the road is capable of AFTER THE MERGE IS OVER, regardless of where the the merge point is.... that is the bottleneck and the zipper merge isn't going to improve that.

The proper metric would be linear feet of congestion. Does the zipper merge effectively decrease the number of linear feet of congestion? Because if an early merge is causing a 5 mile backup and switching it to a zipper merge reduces the backup by 1 mile, that means that anyone exiting in that one mile that the zipper merge reduces the congestion will now no longer experience any congestion, have their travel time cut down, and reduce the congestion for everyone by getting less people in the congested areas.

If you insist on using VPH as your metric, you actually can use that as well.... you just need to make sure you use it at the right spot on the road. The VPH metric will only be increased by measuring before each exit prior to the merge point. The VPH will mostly be increased at the first exit after the congestion would otherwise begin, with smaller increases as you get closer to the zipper merge. Using VPH close to the actual merge point is a meaningless metric, thats not where the zipper merge is designed to increase it.

When people fail to use the road capacity, the length of congestion increases. That is the metric that you are supposed to be using

Good example, thank you.
However - and we discussed that somewhere upstream - such approach is relevant in limited situations only. Zipper concept is pushed mostly for highway traffic. Then in order for all that to be relevant, an exit has to be close enough to the bottleneck; and traffic volume has to be such that an extra half a mile of jam - due to early merge a mile before the bottleneck makes a difference; a bit more traffic - and exit blocked anyway; a bit less - and exit is free anyway.
Probably can be important in dense cities - NYC and LA...

[jemacedo9]

Back when I was in high school, they built a new building and did not expand the parking lot (parking capacity was not a real issue, before the expansion the lot would NEVER come close to filling up). After this, they ran into an issue where at pickup time, it would back up onto the main road and cause all sorts of problems, once it backed up onto it, it'd back up into other intersections and create gridlock. To try to solve this, they converted the parking lanes leading to the pickup area to be one way and "two lanes", and put signs at the beginning of this they put a sign "USE BOTH LANES" and at the end of it they put a sign "MERGE HERE - ALTERNATE LANES". The parents mostly ignored the signs, so they assumed they didn't see them and made additional signs, and installed 2 more of the same signs at the beginning and end, except one on the right side of the road, one in the center between the two lanes, and on on the left side. The parents still ignored the signs... they then decided to station the school officer by the sign to help people alternate between the two lanes... as soon as people were 50 feet past the person, they'd try to move back to the right lane way before the end of it. They added "STAY IN YOUR LANE" signs after the USE BOTH LANE signs and they didn't help. They started laying out cones to keep people in the lanes until the end which seemed effective, but any time someone wasn't out to direct the traffic, almost everyone would inevitably end up in the right lane and it'd backup onto the main road. When I asked parents why they wouldn't use the left lane, they said it wasn't going to make anything any faster, the pickup area only moves so fast, and they needed to be in the right lane for pickup anyways.

The length of the queue wasn't really off by too much... if they got about 10-15 cars into the left lane and people to pull forward and not leave multiple car lengths, the entire queue would be within the school property and it'd prevent the backups onto other roads. After 2 or 3 years of this, they redid the parking lot to add an extra loop to the queue so the cars could remain in a single lane.

So whats the point of this story? Its literally demonstrating the purpose of the zipper merge. So back to the topic:

You'll be surprised how many people on this forum do disagree with VPH metrics.. 
But once you adopt that metric, entire zipper concept falls apart.

Thats because VPH is the WRONG metric of what the zipper concept is trying to solve. The VPH through the merge point is ultimately what the road is capable of AFTER THE MERGE IS OVER, regardless of where the the merge point is.... that is the bottleneck and the zipper merge isn't going to improve that.

The proper metric would be linear feet of congestion. Does the zipper merge effectively decrease the number of linear feet of congestion? Because if an early merge is causing a 5 mile backup and switching it to a zipper merge reduces the backup by 1 mile, that means that anyone exiting in that one mile that the zipper merge reduces the congestion will now no longer experience any congestion, have their travel time cut down, and reduce the congestion for everyone by getting less people in the congested areas.

If you insist on using VPH as your metric, you actually can use that as well.... you just need to make sure you use it at the right spot on the road. The VPH metric will only be increased by measuring before each exit prior to the merge point. The VPH will mostly be increased at the first exit after the congestion would otherwise begin, with smaller increases as you get closer to the zipper merge. Using VPH close to the actual merge point is a meaningless metric, thats not where the zipper merge is designed to increase it.

When people fail to use the road capacity, the length of congestion increases. That is the metric that you are supposed to be using

Good example, thank you.
However - and we discussed that somewhere upstream - such approach is relevant in limited situations only. Zipper concept is pushed mostly for highway traffic. Then in order for all that to be relevant, an exit has to be close enough to the bottleneck; and traffic volume has to be such that an extra half a mile of jam - due to early merge a mile before the bottleneck makes a difference; a bit more traffic - and exit blocked anyway; a bit less - and exit is free anyway.
Probably can be important in dense cities - NYC and LA...

It's not just exits.  For example, if there is a curve on the highway 2 miles before the work zone, which creates a dangerous situation if traffic backs up to just after the curve; and a zipper merge helps prevent backups from reaching that curve, then I'd say the zipper merge is preferable.

This argument of zipper merges are always preferable...or never preferable, is missing the point.  There are situations for one vs the other.  The difference is specific to roadway geometry (curves, exits, etc) and volume.  A lower volume highway and simple geometry, zipper merge is probably useless; early merge is probably preferable. High volume road, complex geometry; zipper merge is preferable assuming all users understand the concept.

[kalvado]

It's not just exits.  For example, if there is a curve on the highway 2 miles before the work zone, which creates a dangerous situation if traffic backs up to just after the curve; and a zipper merge helps prevent backups from reaching that curve, then I'd say the zipper merge is preferable.

And this is an example of bad argument, to the point of getting F for the test just for that thing alone for inability to see bigger picture.
Once there is a backup, there will be a slowdown point. If it falls into the curve, right before or right after  is more or less pure luck. This is more of an arguments towards how speed should be controlled to enable safe stopping even in complex geometry - but nothing else

[UCFKnights]

Back when I was in high school, they built a new building and did not expand the parking lot (parking capacity was not a real issue, before the expansion the lot would NEVER come close to filling up). After this, they ran into an issue where at pickup time, it would back up onto the main road and cause all sorts of problems, once it backed up onto it, it'd back up into other intersections and create gridlock. To try to solve this, they converted the parking lanes leading to the pickup area to be one way and "two lanes", and put signs at the beginning of this they put a sign "USE BOTH LANES" and at the end of it they put a sign "MERGE HERE - ALTERNATE LANES". The parents mostly ignored the signs, so they assumed they didn't see them and made additional signs, and installed 2 more of the same signs at the beginning and end, except one on the right side of the road, one in the center between the two lanes, and on on the left side. The parents still ignored the signs... they then decided to station the school officer by the sign to help people alternate between the two lanes... as soon as people were 50 feet past the person, they'd try to move back to the right lane way before the end of it. They added "STAY IN YOUR LANE" signs after the USE BOTH LANE signs and they didn't help. They started laying out cones to keep people in the lanes until the end which seemed effective, but any time someone wasn't out to direct the traffic, almost everyone would inevitably end up in the right lane and it'd backup onto the main road. When I asked parents why they wouldn't use the left lane, they said it wasn't going to make anything any faster, the pickup area only moves so fast, and they needed to be in the right lane for pickup anyways.

The length of the queue wasn't really off by too much... if they got about 10-15 cars into the left lane and people to pull forward and not leave multiple car lengths, the entire queue would be within the school property and it'd prevent the backups onto other roads. After 2 or 3 years of this, they redid the parking lot to add an extra loop to the queue so the cars could remain in a single lane.

So whats the point of this story? Its literally demonstrating the purpose of the zipper merge. So back to the topic:

You'll be surprised how many people on this forum do disagree with VPH metrics.. 
But once you adopt that metric, entire zipper concept falls apart.

Thats because VPH is the WRONG metric of what the zipper concept is trying to solve. The VPH through the merge point is ultimately what the road is capable of AFTER THE MERGE IS OVER, regardless of where the the merge point is.... that is the bottleneck and the zipper merge isn't going to improve that.

The proper metric would be linear feet of congestion. Does the zipper merge effectively decrease the number of linear feet of congestion? Because if an early merge is causing a 5 mile backup and switching it to a zipper merge reduces the backup by 1 mile, that means that anyone exiting in that one mile that the zipper merge reduces the congestion will now no longer experience any congestion, have their travel time cut down, and reduce the congestion for everyone by getting less people in the congested areas.

If you insist on using VPH as your metric, you actually can use that as well.... you just need to make sure you use it at the right spot on the road. The VPH metric will only be increased by measuring before each exit prior to the merge point. The VPH will mostly be increased at the first exit after the congestion would otherwise begin, with smaller increases as you get closer to the zipper merge. Using VPH close to the actual merge point is a meaningless metric, thats not where the zipper merge is designed to increase it.

When people fail to use the road capacity, the length of congestion increases. That is the metric that you are supposed to be using

Good example, thank you.
However - and we discussed that somewhere upstream - such approach is relevant in limited situations only. Zipper concept is pushed mostly for highway traffic. Then in order for all that to be relevant, an exit has to be close enough to the bottleneck; and traffic volume has to be such that an extra half a mile of jam - due to early merge a mile before the bottleneck makes a difference; a bit more traffic - and exit blocked anyway; a bit less - and exit is free anyway.
Probably can be important in dense cities - NYC and LA...

I can see the same thing in Orlando every day, and when I was in South Florida, there too (although people there were better about zipper merging). And Jacksonville too. Infact, almost anywhere that isn't rural with exits many miles a part. The thing is, just like my school example, where once the traffic backs up to the main road, even though its only a few cars in the main road, you're now blocking all the people who are not even part of the merge, and they have to be added to the queue, which led to it backing up into other intersections, and now creating congestion on other roads. Once you block that first exit, or cause the entry ramp onto the highway to back up onto another road, the growth of the congestion isn't linear. Now you have to add everyone who wants to take that first exit to the congestion, in addition to the upstream problem. Which causes it to block 2 exits back, and so on.

People seriously underestimate how much of a difference small pieces of road and their utilization can make to capacity and congestion.

[kphoger]

The throughput at the bottleneck remains the same either way, right?  All that's really changing with a zipper merge is that, between the two lanes, more storage is accomplished with the zipper merge.  Plus it also discourages the mentality that "this is my lane, dammit, so you can just sit there and wait because you didn't move over when I think you should have."

I've seen a lot of argument as to why a zipper merge might not actually be benificial, but where are the arguments that it's detrimental?

[kalvado]

The throughput at the bottleneck remains the same either way, right?  All that's really changing with a zipper merge is that, between the two lanes, more storage is accomplished with the zipper merge.  Plus it also discourages the mentality that "this is my lane, dammit, so you can just sit there and wait because you didn't move over when I think you should have."

I've seen a lot of argument as to why a zipper merge might not actually be benificial, but where are the arguments that it's detrimental?

Actually zipper slightly  reduces throughput at the bottleneck. Having to do two things - lane change and accelerate, while making sure driver in the other lane actually doing what they are supposed to do - will slow down entry into the limiting point.
And well, legality of zipper approach is quite questionable.

[kalvado]

I can see the same thing in Orlando every day, and when I was in South Florida, there too (although people there were better about zipper merging). And Jacksonville too. Infact, almost anywhere that isn't rural with exits many miles a part. The thing is, just like my school example, where once the traffic backs up to the main road, even though its only a few cars in the main road, you're now blocking all the people who are not even part of the merge, and they have to be added to the queue, which led to it backing up into other intersections, and now creating congestion on other roads. Once you block that first exit, or cause the entry ramp onto the highway to back up onto another road, the growth of the congestion isn't linear. Now you have to add everyone who wants to take that first exit to the congestion, in addition to the upstream problem. Which causes it to block 2 exits back, and so on.

People seriously underestimate how much of a difference small pieces of road and their utilization can make to capacity and congestion.

As discussed upstream, blocking entry ramp may actually be beneficial as it encourages using alternative roads.  Those not using GPS software may miss the message, though..

[kphoger]

The throughput at the bottleneck remains the same either way, right?  All that's really changing with a zipper merge is that, between the two lanes, more storage is accomplished with the zipper merge.  Plus it also discourages the mentality that "this is my lane, dammit, so you can just sit there and wait because you didn't move over when I think you should have."

I've seen a lot of argument as to why a zipper merge might not actually be benificial, but where are the arguments that it's detrimental?

Actually zipper slightly  reduces throughput at the bottleneck. Having to do two things - lane change and accelerate, while making sure driver in the other lane actually doing what they are supposed to do - will slow down entry into the limiting point.
And well, legality of zipper approach is quite questionable.

It's my understanding that, if done correctly, the zipper merge should eliminate the existence of an "other lane".  That is, both lanes are equally responsible for merging and letting in.  That's what taking turns means.

[kalvado]

The throughput at the bottleneck remains the same either way, right?  All that's really changing with a zipper merge is that, between the two lanes, more storage is accomplished with the zipper merge.  Plus it also discourages the mentality that "this is my lane, dammit, so you can just sit there and wait because you didn't move over when I think you should have."

I've seen a lot of argument as to why a zipper merge might not actually be benificial, but where are the arguments that it's detrimental?

Actually zipper slightly  reduces throughput at the bottleneck. Having to do two things - lane change and accelerate, while making sure driver in the other lane actually doing what they are supposed to do - will slow down entry into the limiting point.
And well, legality of zipper approach is quite questionable.

It's my understanding that, if done correctly, the zipper merge should eliminate the existence of an "other lane".  That is, both lanes are equally responsible for merging and letting in.  That's what taking turns means.

What do you mean by "eliminating"? Would bulldozers move in and remove the pavement - or do you mean something else?

[kphoger]

The throughput at the bottleneck remains the same either way, right?  All that's really changing with a zipper merge is that, between the two lanes, more storage is accomplished with the zipper merge.  Plus it also discourages the mentality that "this is my lane, dammit, so you can just sit there and wait because you didn't move over when I think you should have."

I've seen a lot of argument as to why a zipper merge might not actually be benificial, but where are the arguments that it's detrimental?

Actually zipper slightly  reduces throughput at the bottleneck. Having to do two things - lane change and accelerate, while making sure driver in the other lane actually doing what they are supposed to do - will slow down entry into the limiting point.
And well, legality of zipper approach is quite questionable.

It's my understanding that, if done correctly, the zipper merge should eliminate the existence of an "other lane".  That is, both lanes are equally responsible for merging and letting in.  That's what taking turns means.

What do you mean by "eliminating"? Would bulldozers move in and remove the pavement - or do you mean something else?

I meant something else.  To speak of "the other lane" indicates believing there is one lane that drivers should be in and another lane drivers should get out of.  A zipper merge should eliminate this mentality, encouraging both lanes to simply take turns at the taper.  In Australia, this would be defined as the difference between "merging" and "forming one lane".  (cf ACT road rules handbook, page 38, on the difference between the two)

[kalvado]

The throughput at the bottleneck remains the same either way, right?  All that's really changing with a zipper merge is that, between the two lanes, more storage is accomplished with the zipper merge.  Plus it also discourages the mentality that "this is my lane, dammit, so you can just sit there and wait because you didn't move over when I think you should have."

I've seen a lot of argument as to why a zipper merge might not actually be benificial, but where are the arguments that it's detrimental?

Actually zipper slightly  reduces throughput at the bottleneck. Having to do two things - lane change and accelerate, while making sure driver in the other lane actually doing what they are supposed to do - will slow down entry into the limiting point.
And well, legality of zipper approach is quite questionable.

It's my understanding that, if done correctly, the zipper merge should eliminate the existence of an "other lane".  That is, both lanes are equally responsible for merging and letting in.  That's what taking turns means.

What do you mean by "eliminating"? Would bulldozers move in and remove the pavement - or do you mean something else?

I meant something else.  To speak of "the other lane" indicates believing there is one lane that drivers should be in and another lane drivers should get out of.  A zipper merge should eliminate this mentality, encouraging both lanes to simply take turns at the taper.  In Australia, this would be defined as the difference between "merging" and "forming one lane".  (cf ACT road rules handbook, page 38, on the difference between the two)

You may eliminate "my lane" mentality, but you cannot eliminate the fact that there are two (or more) vehicles with intention to proceed into the same spot at the same time. You may implement different rules and laws addressing the situation (one yields, both are equal, courtesy of giving more breathing space to a more needy vehicle - e.g. heavy 18-wheeler having hard time accelerating), but not the fact that there is this lane and that lane - and potential for the accident when this and that do not cooperate.

And legislation may be different between different areas. I may be wrong, but I didn't see any discrepancies in state rules saying that traffic on the main highway has right of way over traffic on ramp. Looks like Australia thinks differently.
But talking about US with "ramp traffic yields" laws -  as we talked upstream, situations where you handpick one set of rules over the other set in each specific situation is a recipe for an accident. We can say that ramp merge is clearly different from construction zone merge - but soon there will be examples of situations in between, causing all sort of problems.

[kphoger]

situations where you handpick one set of rules over the other set in each specific situation is a recipe for an accident. We can say that ramp merge is clearly different from construction zone merge - but soon there will be examples of situations in between, causing all sort of problems.

But this is how things change.  Any new thing is experimented with in certain locations in specific situations, then other places like the idea and try it out.  Eventually, sloppily, and haphazardly, it–whatever it is–becomes mainstream and even finds its way into the MUTCD.  Hardly anything just gets dropped on every highway in the nation at once like a nuke.