Lane drops versus lane ends

[Duke87]

Say you have a section of six lane freeway you want to transition into a section of four lane freeway. This almost always happens at or right after an interchange for fairly logical reasons, although there are two ways of doing it. You can use the interchange ramps to drop the extra pair of lanes:


Or you can continue them past the ramps and then end them:


Which do you prefer?

I, personally, tend to prefer the former. Continuing them and ending them when you have a convenient place to drop them right there just seems... sloppy.
Although, I can see some merit to "continue and end" when the interchange in question isn't a major one. Especially if you're considering a future widening of the four lane part. And indeed, it does seem to be the more common practice in those cases...

[Truvelo]

The first option works if there is a large proportion of traffic leaving at that exit. If not, you tend to find people bombing along the empty right hand lane and then cutting in at the last minute before the exit.

Over here if there isn't much traffic using that exit the lane drop would normally occur before the exit.

[froggie]

The first option works if there is a large proportion of traffic leaving at that exit. If not, you tend to find people bombing along the empty right hand lane and then cutting in at the last minute before the exit.

Agreed.

Over here if there isn't much traffic using that exit the lane drop would normally occur before the exit.

But then you're losing lane balance/lane continuity by dropping it before the exit.  The #2 that Duke posted works better in that regard...

[Chris]

I think it should represent traffic flows. Where do traffic flows change? Right, at interchanges. If a lane drop has to be done, I'd favor the left lane to end, because truck traffic is mostly in the right lanes. I think almost all lane drops in Europe are on the left lane if it isn't at an interchange, but mainly because our trucks are much more strictly keeping right.

The sight of trucks just driving in the middle lanes if they are not directly overtaking someone is really rare in Europe.

[Truvelo]

The problem with trucks being in the 'fast' lane is dealt with layouts like this.

[froggie]

But you're still creating lane inbalance by dropping the lane before the exit...

[deathtopumpkins]

I prefer the lane to exit, rather than just end. Though I've noticed that standard practice around these parts involves neither: Another lane will be added to the right, which will exit only, and then the original right lane will become an optional lane that ends shortly after the exit, such as HERE and HERE.

[Bryant5493]

Both have their benefits, of course.

Exit-only lanes are problematic, because many drivers don't signal their intentions and, other times, people aren't courteous enough to let folks in or out of these lanes. Lanes that end, at least the ones I've seen, aren't marked well -- especially the ones in Georgia. I prefer what Alabama does: have the short dashed lines continue to the end of the lane, rather than having several feet of unmarked pavement. Additionally, I like having the "merge" arrow drawn on the pavement, accompanied by the words "merge left" or "merge right," respectively.


Be well,

Bryant

[Brandon]

Both are fucntionally better than what IDOT (IL) decided to for this on I-55 for several years.  IDOT, in their infinite wisdom, decided to end the right lane on I-55 at milepost 264, one mile before the Weber Road exit (Exit 263).  It seemed to take them several years to understand that either the lane needed to exit at Weber Road (which they finally did until widening I-55) or end the lane after the busy exit.

[akotchi]

My two key factors are volume distribution and context within the freeway system.

My real issue is that neither case is particularly well signed consistently.

[mightyace]

I don't have a preference one way or another and both types exist here in Middle-TN.  They are, IMHO, well signed here.  Yet, people still make quick shifts from exit only lanes and wait until the lane is ending on lane ends.  (They do that on road construction two.)

It almost seems at times that they don't believe that the lane is exiting/ending! 

[Alps]

Adding a lane is best done at the interchange ramp itself, not prior.  That way it's a smooth entry and not a merge condition.  Dropping a lane is best done after the ramp, not at it, because the turbulence in the interchange area is already lowering capacity.  By separating the two capacity-reducing areas from each other, you're not compounding the effect, and you get a few hundred VPH extra out of the deal.

[deathtopumpkins]

It almost seems at times that they don't believe that the lane is exiting/ending! 

I've found this often to actually be the case, at least around here, particularly with lane closures for road work. Often the signs will be put up hours before the work begins or left up overnight even when the lane is not still closed, so instead of joining all the slow cars who get over as soon as they see a lane ends sign, people will zip around them all in the lane thats supposed to be closed.

[roadfro]

Adding a lane is best done at the interchange ramp itself, not prior.  That way it's a smooth entry and not a merge condition.  Dropping a lane is best done after the ramp, not at it, because the turbulence in the interchange area is already lowering capacity.  By separating the two capacity-reducing areas from each other, you're not compounding the effect, and you get a few hundred VPH extra out of the deal.

I agree with your response on added lanes, but I think the dropped lane is a different matter.  It really depends on what the volume, origin-destination paths, and the distance from previous ramps are upstream from the exit ramp at the interchange.  If the exit ramp leads to a major highway or traffic generator and there is adequate spacing from the previous ramp, it could be more beneficial to have the lane drop at the ramp.

[njroadhorse]

Adding a lane is best done at the interchange ramp itself, not prior.  That way it's a smooth entry and not a merge condition.  Dropping a lane is best done after the ramp, not at it, because the turbulence in the interchange area is already lowering capacity.  By separating the two capacity-reducing areas from each other, you're not compounding the effect, and you get a few hundred VPH extra out of the deal.

I agree with your response on added lanes, but I think the dropped lane is a different matter.  It really depends on what the volume, origin-destination paths, and the distance from previous ramps are upstream from the exit ramp at the interchange.  If the exit ramp leads to a major highway or traffic generator and there is adequate spacing from the previous ramp, it could be more beneficial to have the lane drop at the ramp.

Or do it after the ramp, once the on-ramp has merged on.  I've seen this done a few times, mostly in Maryland.

[haljackey]

I would prefer to see a hybrid:



(Assuming right hand drive here)

That way drivers entering don't have to worry about merging and those in the ending right lane don't have to worry about exiting.

[J N Winkler]

Something like this has been tried at various exit ramps on I-27 in Lubbock, Texas:

http://maps.google.com/maps?q=Lubbock,+Texas&ie=UTF8&hq=&hnear=Lubbock,+Texas&ll=33.543541,-101.844469&spn=0.001151,0.003433&t=k&z=19

Essentially, instead of the gore being bounded by solid lines on both sides, the mainline side is an ordinary traffic lane stripe.  The intention seems to be to give drivers space to merge back onto the freeway if they decide they do not want to take the exit and drive across the gore.  I don't particularly like this approach since it makes the gore difficult to locate visually, especially at night and in rain.