New MUTCD announced

[jeffandnicole]

So use the available lanes till they end, because the constriction point won't change because traffic vacates an open lane early.

Try again with a better picture:


Zone A: incoming free-flowing traffic slows down due to congestion ahead and merges into a single file, proceding like that into Area B
....
....
Area D: outgoing traffic is now not restricted, and accelerates to a highway  free-flow speed

Can you describe zone B and C?  assume no lane jumpers, everyone merges.

PS pardon my crazy image editing, don't shoot the piano player

Area C has no actual work going on?  Just a closed lane?

Then traffic after it merges in will start going faster.  The length of the jammed traffic will remain the same length if they merge together at the lane's actual end, or at some random point 1/2 mile prior to the lane ending.  If traffic is jammed a mile, it's going to be jammed a mile regardless.  The difference will be the area of highway that is jammed will be 1/2 mile closer to the lane ending point.

Thus, there's no benefit to merging 1/2 mile ahead.

If there's work going on in that Area C work zone - flashing lights, work near the skip line, large trucks turning, etc, that will get people's attention, just like cars in a crash.  Traffic will remain slow until they pass by the attention-causing stuff, at which point it will start moving faster.  But again, it won't matter if traffic merges at the end of the lane or a random point prior to the end of the lane; it's going to remain jammed until they get by the issue.

[kphoger]

Zone A: incoming free-flowing traffic slows down due to congestion ahead and merges into a single file, proceding like that into Area B
....
....
Area D: outgoing traffic is now not restricted, and accelerates to a highway  free-flow speed

Can you describe zone B and C?  assume no lane jumpers, everyone merges.

Zone B:  You said everyone merged into single file back in Zone A, so it's now single-file traffic in Zone B.  Except that, because you shifted the merge point a half-mile upstream from the orange cones, that single file might now extend farther upstream—depending on how heavy traffic is.

Zone C:  It depends.  If there's not much going on in the work zone, then traffic recovers from the slack-action of merging;  I suppose this is what you called "flow stabilization".  However, if there is a lane shift, or a piece of heavy equipment near the lane of traffic, or dust in the air, or whatever—then traffic will slow down again anyway;  this is what J&N was talking about.  So, by putting the theoretical choke point upstream of the physical choke point, either you might have helped stabilize the flow ahead of the one-lane stretch, or else you've introduced a second slow-down point that wouldn't otherwise have existed.

[kalvado]

So use the available lanes till they end, because the constriction point won't change because traffic vacates an open lane early.

Try again with a better picture:


Zone A: incoming free-flowing traffic slows down due to congestion ahead and merges into a single file, proceding like that into Area B
....
....
Area D: outgoing traffic is now not restricted, and accelerates to a highway  free-flow speed

Can you describe zone B and C?  assume no lane jumpers, everyone merges.

PS pardon my crazy image editing, don't shoot the piano player

Area C has no actual work going on?  Just a closed lane?

Then traffic after it merges in will start going faster.  The length of the jammed traffic will remain the same length if they merge together at the lane's actual end, or at some random point 1/2 mile prior to the lane ending.  If traffic is jammed a mile, it's going to be jammed a mile regardless.  The difference will be the area of highway that is jammed will be 1/2 mile closer to the lane ending point.

Thus, there's no benefit to merging 1/2 mile ahead.

If there's work going on in that Area C work zone - flashing lights, work near the skip line, large trucks turning, etc, that will get people's attention, just like cars in a crash.  Traffic will remain slow until they pass by the attention-causing stuff, at which point it will start moving faster.  But again, it won't matter if traffic merges at the end of the lane or a random point prior to the end of the lane; it's going to remain jammed until they get by the issue.

good!
I would say, though, that even with no ongoing work traffic would accelerate after passing into the narrows, between B and C
So my description is:
Zone B: slow flow in congested mode
Zone C: varies, depending on condition, but generally faster than zone B

are you still with me?

[jeffandnicole]

So use the available lanes till they end, because the constriction point won't change because traffic vacates an open lane early.

Try again with a better picture:


Zone A: incoming free-flowing traffic slows down due to congestion ahead and merges into a single file, proceding like that into Area B
....
....
Area D: outgoing traffic is now not restricted, and accelerates to a highway  free-flow speed

Can you describe zone B and C?  assume no lane jumpers, everyone merges.

PS pardon my crazy image editing, don't shoot the piano player

Area C has no actual work going on?  Just a closed lane?

Then traffic after it merges in will start going faster.  The length of the jammed traffic will remain the same length if they merge together at the lane's actual end, or at some random point 1/2 mile prior to the lane ending.  If traffic is jammed a mile, it's going to be jammed a mile regardless.  The difference will be the area of highway that is jammed will be 1/2 mile closer to the lane ending point.

Thus, there's no benefit to merging 1/2 mile ahead.

If there's work going on in that Area C work zone - flashing lights, work near the skip line, large trucks turning, etc, that will get people's attention, just like cars in a crash.  Traffic will remain slow until they pass by the attention-causing stuff, at which point it will start moving faster.  But again, it won't matter if traffic merges at the end of the lane or a random point prior to the end of the lane; it's going to remain jammed until they get by the issue.

good!
I would say, though, that even with no ongoing work traffic would accelerate after passing into the narrows, between B and C
So my description is:
Zone B: slow flow in congested mode
Zone C: varies, depending on condition, but generally faster than zone B

are you still with me?

Nope, not with you.  I'm providing actual examples of why traffic slows down; examples that can be viewed and witnessed.  You state an opinion by using "I will say" without any justification for it.

[kalvado]

Nope, not with you.  I'm providing actual examples of why traffic slows down; examples that can be viewed and witnessed.  You state an opinion by using "I will say" without any justification for it.

Since this is a critical point of discussion, lets slow down. There is no traffic to push us from the back, we can go slowly.
Unfortunately both of us would have to go by hand-waving mostly. I tried to look for some aerials of real life traffic jams, but didn't find good ones. 

If there's work going on in that Area C work zone - flashing lights, work near the skip line, large trucks turning, etc, that will get people's attention, just like cars in a crash.  Traffic will remain slow until they pass by the attention-causing stuff, at which point it will start moving faster.  But again, it won't matter if traffic merges at the end of the lane or a random point prior to the end of the lane; it's going to remain jammed until they get by the issue.

Absolutely true. moreover - things depend a lot on where in the narrows that happens and what happens. Thats why I  wanted to postpone this iteration. 
Let me formulate it so - if there is more than one bottleneck, one of them would be the limiting area. If it is located upstream, there will be no problems downstream.  If downstream point is worse, a second queue of congestion will grow, eventually coalescing  with upstream one.

Does that make any sense? 

[lordsutch]

The advanced merge concept laid out here does either one of two things: (a) it moves the zipper merge from the actual lane closure to the point where some sign says to merge, which just lengthens the single-file work zone by the length of Area B, or (b) it just allows drivers to continue to jockey lanes through the length of Area B, absent active enforcement, recreating multiple merge points and getting rid of the zipper merge in practice.

All you've done, even if everyone complies with the instruction to merge 1/2 mile ahead of the lane closure, is extend the slower single file zone by 1/2 mile, and you've reduced the storage space for traffic upstream of the closure by 1/2 a lane-mile. But as noted above compliance will not be 100%, so now you're back to an unholy mix of some drivers zipper merging and some drivers trying to jockey for position before the actual lane closure like they're at Talladega, and everyone being pissed off.

[NoGoodNamesAvailable]

I'm not going to add anything to the zipper merge debate except that I'm an avid zipper merger myself. But if I was an engineer I would never call for this signage. It's too risky. Especially the regulatory "stay in lane to merge point" sign. Besides the fact that I don't think any law supports that... when somebody is drowsy/drunk/distracted and plows into the work zone at 70 mph, possibly killing workers, now the engineer of record is going to be implicated in the lawsuit.

Zipper merging makes sense in heavy traffic. But it should be a judgement call by the drivers. In free flow traffic conditions, a sign telling people to stay in the lane that's ending is not safe.

[vdeane]

I'm not going to add anything to the zipper merge debate except that I'm an avid zipper merger myself. But if I was an engineer I would never call for this signage. It's too risky. Especially the regulatory "stay in lane to merge point" sign. Besides the fact that I don't think any law supports that... when somebody is drowsy/drunk/distracted and plows into the work zone at 70 mph, possibly killing workers, now the engineer of record is going to be implicated in the lawsuit.

Zipper merging makes sense in heavy traffic. But it should be a judgement call by the drivers. In free flow traffic conditions, a sign telling people to stay in the lane that's ending is not safe.

Agreed.  Not to mention that it could cause other issues.  I prefer to zipper merge myself, however it depends on traffic.  If traffic is already gnarled, I zipper merge - no need to merge early, for all the reasons people traditionally advocate for zipper merges for.  If the lane I need to merge into is clear and will remain so, again I move over at the merge point - no reason not to.  The exception is if the lane I need to merge into is clear now, but cars are approaching and I have reason to believe it might not be later.  Then, I'll merge immediately, since it's less of a strain of traffic for me to merge into the open space rather than force someone to let me in.

[Scott5114]

I understand the theory behind zipper merge. But I don't trust any other driver to understand it enough to actually let me practice it. So I merge early.

[jeffandnicole]

I understand the theory behind zipper merge. But I don't trust any other driver to understand it enough to actually let me practice it. So I merge early.

Go to a crowded metro area.  It's a daily routine.  People in more rural areas tend to be 'nicer' and think they should get over earlier.

[jeffandnicole]

I'm not going to add anything to the zipper merge debate except that I'm an avid zipper merger myself. But if I was an engineer I would never call for this signage. It's too risky. Especially the regulatory "stay in lane to merge point" sign. Besides the fact that I don't think any law supports that... when somebody is drowsy/drunk/distracted and plows into the work zone at 70 mph, possibly killing workers, now the engineer of record is going to be implicated in the lawsuit.

Zipper merging makes sense in heavy traffic. But it should be a judgement call by the drivers. In free flow traffic conditions, a sign telling people to stay in the lane that's ending is not safe.

So I agree with some of this:  That sign is really meant when there is congestion.  But it should be stated that way.  When traffic is free-flowing, like Ms. Deane mentioned, I'll merge early also when there's no congestion.

However, referencing the drunk:  The defense trying to claim a drunk understood the sign to stay in the closed lane is going to be immediate refuted that he should have also understood lane ending signage, cones, barrels, a large flashing arrow, safety trucks, and other implements of construction.  Courts don't really give drunks much sympathy in their actions, especially when there's a death.  And the engineer is going to have the MUTCD as his document that he followed standard accepted protocol.

[jay8g]

Back to the new MUTCD...

One of the strangest changes I've noticed so far is the decision to limit the standard R3-5 signs (right/left turn only) to overhead use only. It seems that this was intended to restrict the usage that has become common in Seattle (among other places) where the R3-5 is used instead of the "RIGHT LANE MUST TURN RIGHT" sign or a R3-8-type sign showing all lanes, but they also restricted the extremely common case where the sign is used for locations where all traffic must turn, and instead reintroduced the all-text "RIGHT TURN ONLY" sign as the R4-21 (alongside a strange "ALL TRAFFIC" sign as the R4-20, which appears to be intended for use at areas with splitter islands).


I really don't understand the point of this change, and I expect that this is something that most agencies will ignore or just not understand. I wouldn't be surprised to see this get rolled back in a future edition.

On a side note, it's interesting how old-fashioned some of the "new" signs in this MUTCD look -- particularly the R4-21, which is a blast from the past from when text-only signs were the standard, but also things like the new R3-19 "LANE FOR LEFT TURN ONLY" sign.

[Scott5114]

I understand the theory behind zipper merge. But I don't trust any other driver to understand it enough to actually let me practice it. So I merge early.

Go to a crowded metro area.  It's a daily routine.  People in more rural areas tend to be 'nicer' and think they should get over earlier.

In metro areas, someone invariably decides to ride alongside my rear fender at the exact speed that I'm going so I can't get over. Because how can you ever feel like a real man if you let someone in a sedan merge in front of your pickup truck?

[mgk920]

I understand the theory behind zipper merge. But I don't trust any other driver to understand it enough to actually let me practice it. So I merge early.

Go to a crowded metro area.  It's a daily routine.  People in more rural areas tend to be 'nicer' and think they should get over earlier.

In metro areas, someone invariably decides to ride alongside my rear fender at the exact speed that I'm going so I can't get over. Because how can you ever feel like a real man if you let someone in a sedan merge in front of your pickup truck?

Sad to say it, but that is the time when you sloooooely atart to 'bull' your way in.

Mike

[ran4sh]

My "real data" against the zipper merge is that its proponents want to apply it to all merge types, when it's only really beneficial for low-speed congested merges. At high speed it is unsafe, and with no congestion the zipper merge causes unnecessary congestion of its own.

[Rothman]

My "real data" against the zipper merge is that its proponents want to apply it to all merge types, when it's only really beneficial for low-speed congested merges. At high speed it is unsafe, and with no congestion the zipper merge causes unnecessary congestion of its own.

Wut.

At high speed it is unsafe?  At high speed, there would be little congestion and all you do is adjust your speed to merge in.

If there's no congestion, then there's no congestion.  There's no car to your left, so you just move in.

[ran4sh]

The key phrase was "its proponents want to apply it to all merge types".

[jeffandnicole]

I understand the theory behind zipper merge. But I don't trust any other driver to understand it enough to actually let me practice it. So I merge early.

Go to a crowded metro area.  It's a daily routine.  People in more rural areas tend to be 'nicer' and think they should get over earlier.

In metro areas, someone invariably decides to ride alongside my rear fender at the exact speed that I'm going so I can't get over. Because how can you ever feel like a real man if you let someone in a sedan merge in front of your pickup truck?

Sad to say it, but that is the time when you sloooooely atart to 'bull' your way in.

Mike

Maybe I don't even think about it when I merge in. Been doing it every workday on the office for 20+ years.

[kphoger]

But if I was an engineer I would never call for this signage. It's too risky. Especially the regulatory "stay in lane to merge point" sign. Besides the fact that I don't think any law supports that...

If I'm in the right lane, and there's a slow-moving truck in front of me, then why should I be prohibited from moving into the left lane and passing it?

If I'm in the left lane, and a faster car is approaching me from the rear, then why should I be prohibited from moving into the right lane and letting it pass me?

Yeah, that regulatory sign is bad.

[SEWIGuy]

I understand the theory behind zipper merge. But I don't trust any other driver to understand it enough to actually let me practice it. So I merge early.

Go to a crowded metro area.  It's a daily routine.  People in more rural areas tend to be 'nicer' and think they should get over earlier.

In metro areas, someone invariably decides to ride alongside my rear fender at the exact speed that I'm going so I can't get over. Because how can you ever feel like a real man if you let someone in a sedan merge in front of your pickup truck?

I never merge early and am usually let in by somebody.

[kalvado]

I understand the theory behind zipper merge. But I don't trust any other driver to understand it enough to actually let me practice it. So I merge early.

Go to a crowded metro area.  It's a daily routine.  People in more rural areas tend to be 'nicer' and think they should get over earlier.

In metro areas, someone invariably decides to ride alongside my rear fender at the exact speed that I'm going so I can't get over. Because how can you ever feel like a real man if you let someone in a sedan merge in front of your pickup truck?

I never merge early and am usually let in by somebody.

Just wonder what you do if you're in a long lane, not in the short one?

[SEWIGuy]

I understand the theory behind zipper merge. But I don't trust any other driver to understand it enough to actually let me practice it. So I merge early.

Go to a crowded metro area.  It's a daily routine.  People in more rural areas tend to be 'nicer' and think they should get over earlier.

In metro areas, someone invariably decides to ride alongside my rear fender at the exact speed that I'm going so I can't get over. Because how can you ever feel like a real man if you let someone in a sedan merge in front of your pickup truck?

I never merge early and am usually let in by somebody.

Just wonder what you do if you're in a long lane, not in the short one?

You mean if I let people in?  Yeah, usually I do. In the grand scheme of things it doesn't cost me a lot of time.

Or do you mean, do I leave the long lane for the short one and drive up? Depending on the distance, sure.

[jakeroot]

I think people intuitively understand the concept of "take turns", even if we aren't good about it in-practice.

I see some other roadgeeks (not necessarily here, mostly Facebook groups) that argue failure to merge "early" is what causes backups at the merge point. I don't see how any rational person can come to that conclusion when the clear problem is just constrained capacity from reducing two lanes to one. Going from two lanes to one is fine, if the number of cars in the two lane section is below or at the capacity of the single lane of traffic. But during times of heavy traffic, when even the two lane section is maxing out its capacity, obviously shit is gonna break down when you then force everyone into one lane, no matter where you merge. So, just make it simple and merge where the lanes actually merge.

[kphoger]

I see some other roadgeeks that argue failure to merge "early" is what causes backups at the merge point. I don't see how any rational person can come to that conclusion when the clear problem is just constrained capacity from reducing two lanes to one.

If a late-merger comes to a near-stop (no matter if it's at the orange cones or a half-mile back, but it's typically at the orange cones) and then attempts to merge at 6 mph... then yes, that will cause a problem.  But if people are actually taking turns, or if all traffic is already doing 6 mph anyway, then no problem.

[vdeane]

I think people intuitively understand the concept of "take turns", even if we aren't good about it in-practice.

I see some other roadgeeks (not necessarily here, mostly Facebook groups) that argue failure to merge "early" is what causes backups at the merge point. I don't see how any rational person can come to that conclusion when the clear problem is just constrained capacity from reducing two lanes to one. Going from two lanes to one is fine, if the number of cars in the two lane section is below or at the capacity of the single lane of traffic. But during times of heavy traffic, when even the two lane section is maxing out its capacity, obviously shit is gonna break down when you then force everyone into one lane, no matter where you merge. So, just make it simple and merge where the lanes actually merge.

I think a lot of that comes down to "I merged in, my blood is pumping, I'm ready to go, why should I have to stay stuck in slow traffic to let the late mergers in?".