Best lane drop configuration?

[webny99]

What are your thoughts on how a lane drop near an interchange should be configured? Should it be the same everywhere, or vary based on the situation? Feel free to take the poll and reply to share your thoughts.

[Hobart]

In my opinion, if the main highway is going to drop from 3 lanes in a direction to 2 in a direction after the interchange, the DOT might as well go whole hog and make one of the lanes "exit only" to drop it, instead of ending the lane. I really don't see much of a difference by doing this, especially if it's a parclo with the ramps after the interchange.

However, this isn't the be all end all, and different things need different configurations to be effective!

[Hunty2022]

Right lane becomes an exit only lane. If the exit is on the left side, then the left lane becomes an exit only lane.

[Flint1979]

I like the right lane becoming an exit only lane. I-475 on the south end and I-675 on both ends in Michigan are like that. On the north end of I-475 the left lane ends and the on ramp from exit 126 becomes the exit for I-475.

[jeffandnicole]

Any lane ends should be on the left side, to encourage keeping right except to pass.

I wonder why the poll questions don't have the same options for the right as the left though.

[Henry]

I like it when the right lane is exit only.

[wriddle082]

I also like it when the right lane becomes exit only, but my time in CA has reminded me of what the worst way is: inner lane merges!  CalTrans seems to really like those in the Bay Area, and they're dangerous when you're not used to them.

[webny99]

Any lane ends should be on the left side, to encourage keeping right except to pass.

Definitely agreed on that.

I wonder why the poll questions don't have the same options for the right as the left though.

It does, except for the first choice. That's because the exit being on the right is implied, so only the right lane can become exit only in that scenario.

[wxfree]

In urban areas, I like right lanes that run from entrance to exit.  They're not meant for through traffic, but are weaving lanes.  In principle, I think a lane should end with a merge.  I don't think a long-running lane, that didn't begin at the last entrance ramp, should become exit only.  That requires a lane change, which I see as different from a lane merge.  When you reach the end, you aren't forced off the road, but are able to continue ahead with the need to merge.  As for whether it's right or left, I can see both sides.  At the new Ranger Hill portion of I-20 in Texas, which otherwise has two lanes, the right lane from the entrance ramp from SH 16 ends, and around the same place an additional left lane begins for passing traffic.  This provides three lanes up the hill, for slow trucks, faster trucks, and cars that aren't slowed down by the incline.  In that case, the additional left lane should end.  Ending left lanes generally does seem to remind people of the "keep right" rule.  I like how the stripes push traffic right where a passing lane is added to a two-lane road, so that using the left lane is a lane change rather than the default straight ahead movement.  However, that's a somewhat unnatural movement, because going straight ahead is into the left lane while going to the right lane involves a displacement.  When a lane ends on a multi-lane road, if it's the left lane that ends, that's really the next-to-left lane ending, because the left, center, lane can't end.  There will still be a lane beside the center of the road.  Right lanes, edge lanes, ending is a more natural movement.

[kphoger]

I voted "Right lane ends immediatey after the exit".  Can anyone please explain to me what possible drawback there is to that, compared to "Right lane becomes an exit only lane"?

[webny99]

I voted "Right lane ends immediatey after the exit".  Can anyone please explain to me what possible drawback there is to that, compared to "Right lane becomes an exit only lane"?

I do prefer through lanes to end on the left, but setting that aside, the only reason I can think of is that right lane ending immediately after the exit would allow an additional lane of traffic to flow up to that point, which in turn increases the density of traffic in the right lane approaching the entrance ramp, which in turn makes it harder for entering traffic to find a gap to merge into.

[1995hoo]

I didn't vote in the poll because I don't have a firm opinion. At the interchange nearest to where I live (the Capital Beltway at Van Dorn Street, Exit 173), on the Outer Loop the right lane, which is a thru lane coming from west of Springfield, becomes exit-only and as a general matter I find that layout logical there (despite the number of people who feel "Exit Only" doesn't mean what it says and who will try to use the exit lane as a passing lane), but part of that opinion is no doubt because the thru lanes of I-95 enter to the left a short distance west of there and I thus see a benefit to not having the long-distance traffic's lanes end. So I don't really think there is one overall "best" design because there are so many variables.

With that said, I always liked the way the New Jersey Turnpike signed the lane drop at Exit 4 prior to their replacement of the classic signs:

[jakeroot]

I voted "Right lane ends immediatey after the exit".  Can anyone please explain to me what possible drawback there is to that, compared to "Right lane becomes an exit only lane"?

I do prefer through lanes to end on the left, but setting that aside, the only reason I can think of is that right lane ending immediately after the exit would allow an additional lane of traffic to flow up to that point, which in turn increases the density of traffic in the right lane approaching the entrance ramp, which in turn makes it harder for entering traffic to find a gap to merge into.

Ironically, I think the issue of finding a gap may actually be exacerbated by using an exit-only lane. When a lane ends, that creates a natural merge point; regardless of your view on "late merging", being able to merge at a taper, after the exit ramp, is likely safer (and better for congestion) than potentially having to stop and merge at a more-or-less random point prior to the exit ramp.

Prior to widening, the double-lane on-ramp from I-705 to northbound I-5 in Tacoma, Washington fed into a single exit-only lane that went on for about a mile. During peak traffic hours, the through lanes would frequently back up, and traffic coming in from I-705 would not have a clear place to merge. In practice, traffic would immediately merge as the ramp joined I-5; the left lane would merge into through traffic (a makeshift center lane merge, if you will), with traffic not destined for I-5 using the right lane to bypass them and continuing down the exit only lane. Still, anyone who didn't merge right there would have to merge...somewhere. And they frequently found random places along the exit-only lane to do so. But it meant that they had to nearly stop, forcing traffic behind to drive into the shoulder to bypass them. Having the lane continue beyond the exit and merge would theoretically prevent this, though only if traffic actually used it, of course.

[Ted$8roadFan]

The only problem I have with the right lane ending at an exit is that drivers unfamiliar with the area will often suddenly cut over into the remaining lanes when they realize they're about to leave without meaning to, especially on high volume highways. With the lane drop and proper signage there is more time.

[webny99]

I voted "Right lane ends immediatey after the exit".  Can anyone please explain to me what possible drawback there is to that, compared to "Right lane becomes an exit only lane"?

I do prefer through lanes to end on the left, but setting that aside, the only reason I can think of is that right lane ending immediately after the exit would allow an additional lane of traffic to flow up to that point, which in turn increases the density of traffic in the right lane approaching the entrance ramp, which in turn makes it harder for entering traffic to find a gap to merge into.

Ironically, I think the issue of finding a gap may actually be exacerbated by using an exit-only lane. When a lane ends, that creates a natural merge point; regardless of your view on "late merging", being able to merge at a taper, after the exit ramp, is likely safer (and better for congestion) than potentially having to stop and merge at a more-or-less random point prior to the exit ramp.

I see your point, but it depends on so many different factors that what's best in one location might not be in another. If there's enough traffic exiting via the exit only lane that it's equal speed or slower than the thru lanes, thru traffic would have no incentive to stay in that lane and would instead try to get over early rather than late.

In your example, the on-ramp that would have more trouble finding a gap to merge in this situation would actually be the next one - the one coming in from E Bay St. But that's irrelevant here because that ramp also has its own auxiliary lane.

[jakeroot]

I voted "Right lane ends immediatey after the exit".  Can anyone please explain to me what possible drawback there is to that, compared to "Right lane becomes an exit only lane"?

I do prefer through lanes to end on the left, but setting that aside, the only reason I can think of is that right lane ending immediately after the exit would allow an additional lane of traffic to flow up to that point, which in turn increases the density of traffic in the right lane approaching the entrance ramp, which in turn makes it harder for entering traffic to find a gap to merge into.

Ironically, I think the issue of finding a gap may actually be exacerbated by using an exit-only lane. When a lane ends, that creates a natural merge point; regardless of your view on "late merging", being able to merge at a taper, after the exit ramp, is likely safer (and better for congestion) than potentially having to stop and merge at a more-or-less random point prior to the exit ramp.

I see your point, but it depends on so many different factors that what's best in one location might not be in another. If there's enough traffic exiting via the exit only lane that it's equal speed or slower than the thru lanes, thru traffic would have no incentive to stay in that lane and would instead try to get over early rather than late.

Right, always situationally-dependent. But overall, I would agree with kphoger that there is rarely a situation where lane-drop post-exit actually does any harm. This one in Fife, WA (now gone) was almost never used; the right lane was a de facto exit-only lane, and the previous on-ramp was over 4 miles up the freeway. Still, the few that did use it at least had the option. Now, you have to get off, and I don't see how that's any better.

The main advantage to an exit-only lane, as I see it, is that you can implement a shared exit lane adjacent to it.

[ilpt4u]

Depends. On a roadway with an ongoing 4 to 6 lane years-long project, temp lane ends to the area the new lane is being added (left if inside, right if outside). That way the transition can be easily continued onward as the project continues

Permanent configuration, make it an Exit Only to end the lane

[epzik8]

The most efficient way to do this is for the right lane to become the exit lane.

[sprjus4]

Depends. On a roadway with an ongoing 4 to 6 lane years-long project, temp lane ends to the area the new lane is being added (left if inside, right if outside). That way the transition can be easily continued onward as the project continues

Permanent configuration, make it an Exit Only to end the lane

Interestingly enough, I-64 east of Richmond has undergone a few widening projects, and they've  ended differently. Segments 1-3 between Williamsburg and Newport News, now ends west of VA-199 (MM 234) in Williamsburg with the left lane dropping and a "stub"  in the pavement for future extension.

Outside of Richmond, however, at MM 205 heading eastbound, the right lane becomes exit only and the lanes do a slight "shift"  since the widening occurred in the median, and the left two lanes have to transition back to the original pavement. A slight "stub"  exists at the lane shift though for future left side widening. I feel like one of the disadvantages is it encourages all through traffic to still clog the left two lanes, especially a mile or so before the exit only where the signs read "Exit Only" , but generally I've noticed traffic uses all three lanes before merging near the end.

It's all temporary, however, since VDOT is planning to widen between MM 205 and 216 beginning later this year, and will simply make the existing 3 lanes continuous with a standard deceleration lane for Exit 205.

[Bitmapped]

To me, it depends on context. I prefer lane drops to avoid requiring through traffic from having to do lane changes. If you've added a new lane on the lane, drop it on the left; if it was added on the right, drop it on the right.

I generally prefer dropping on the right as an Exit Only lane, but if the lane being dropped was just added and is a heavy merge situation, I can see benefit in ending it after the offramp but before the onramp. I-77 does this at Exit 102 in Charleston, WV: https://goo.gl/maps/G6fjoxpV9dSmkEVw8

[webny99]

The poll results are a bit surprising so far. I was expecting the options to roughly decrease in popularity, but it's a blowout in favor of "right lane exit only".

I also thought "left lane ends beyond the interchange" would be more popular. IMO this configuration is the least backup-prone, especially approaching the interchange. Even if it is a bit slow at the merge point, it does not affect exiting traffic, which isn't the case when the right lane is exit only because faster traffic tends to use the right lane anyways and weave with exiting traffic if there's a slowdown.

[vdeane]

I think it's because there's a certain elegance to the lane ending as an exit only lane, even if the other configurations have their own advantages.  In fact, I can think of a scenario where using an exit only lane would be undesirable: when there's plans to continue the widening further, since this would complicate the extension.

[Hobart]

I think it's because there's a certain elegance to the lane ending as an exit only lane, even if the other configurations have their own advantages.  In fact, I can think of a scenario where using an exit only lane would be undesirable: when there's plans to continue the widening further, since this would complicate the extension.

Depending on execution, these complexities can be minimized. If the road has an extra wide shoulder with the anticipation of restriping it, all you need to do is some sandblasting of existing lines, and restriping to either push the exit-over lane over to the right, or just make it a non exit-only lane where real estate is limited.

[paulthemapguy]

What we have here is a WEAVING SEGMENT!  This is how the Highway Capacity Manual refers to it.

While I like the neatness of having the acceleration lane blend into the exit only lane, the responsible course of action is to prioritize safety above all else.  The safest consideration, in my opinion, is to provide as much space for the merges as possible.  The onramp's acceleration should continue beyond the exit if there's not enough roadway length to facilitate the weave. If the space between the onramp and the offramp is less than sufficient to keep traffic moving through the weave, extend the acceleration lane downstream of the offramp.  If there's plenty of space to keep traffic moving through the weave, do the "neat" thing and make the added lane into an exit only lane.  I'd ballpark "enough space" around 1/2 mile, but it varies based on roadway parameters.

[kphoger]

I'd ballpark "enough space" around 1/2 mile, but it varies based on roadway parameters.

I'm OK with ½ mile between gore points, but ½ mile between cross streets is a little tight, in my opnion.

Cross-posting from a now-locked thread:

In my experience, weaving is a problem with exits that are spaced a half-mile apart, not exits that are spaced one mile apart.

For example, here's an example of mile-spaced exits with no accel/decel lane, near my office, and it has zero weaving issues.

And here's another one with an accel/decel lane, closer to home, and it also has zero weaving issues even though both exits are fairly busy.

On the other hand, here's an example of half-mile-spaced exits along my commute route, and there are indeed occasional weaving issues–but pretty much only during rush hour.