Interchanges with no stoplights

[Hunty2022]

I-81 @ US/VA-211 in New Market. Controlled with Stop Signs on the ramps.

US 250 @ Rugby Avenue in Charlottesville. Controlled with Stops on ramps + one Yield (US 250 WB to Rugby Ave WB)

VA-7 @ VA-7 BUS in Round Hill. Controlled with Stops + Yields on the ramps.

[1995hoo]

How about this thing in the City of Alexandria between I-95/495 and US-1? There are traffic lights to the south at Fort Hunt Road and to the north at Franklin Street, but neither are part of the interchange. (I-495 in Virginia has two other interchanges with no traffic lights that link to arterials, but they're just ordinary old cloverleaf designs.)

There's a lot going on here. This is a pretty cool design considering the space constraints, but it definitely has more than four ramps (and no at grades, as I see you have noted).

Depending on one's point of view, one could argue that it has only five ramps because most of those start out as single ramps and then split, some of them multiple times. Once you learn what goes where and how to read the signs, it becomes a fairly simple and effective design because one of the main goals was to eliminate weave areas on both roads.

—One from SB US-1 that splits to serve both directions of the Beltway (this one splits three times and serves both the "Local" and "Thru" carriageways in both directions)
—One from the Beltway's outer loop that splits to serve both directions of US-1
—One from NB US-1 that splits to serve both directions of the Beltway  (this one splits twice; it serves both carriageways on the Inner Loop but only the "Local" one on the Outer Loop)
—One from the Beltway's Inner Loop that splits to serve NB US-1 and Church Street (connecting to Washington Street, which becomes the GW Parkway to the south)
—One from the Beltway's Inner Loop that serves SB US-1 (perhaps the most conventional ramp there, a simple loop-around)

[webny99]

How about this thing in the City of Alexandria between I-95/495 and US-1? There are traffic lights to the south at Fort Hunt Road and to the north at Franklin Street, but neither are part of the interchange. (I-495 in Virginia has two other interchanges with no traffic lights that link to arterials, but they're just ordinary old cloverleaf designs.)

There's a lot going on here. This is a pretty cool design considering the space constraints, but it definitely has more than four ramps (and no at grades, as I see you have noted).

Depending on one's point of view, one could argue that it has only five ramps because most of those start out as single ramps and then split, some of them multiple times. Once you learn what goes where and how to read the signs, it becomes a fairly simple and effective design because one of the main goals was to eliminate weave areas on both roads.

I would probably consider the splits to be separate ramps, but still, only five mainline exit ramps for the entire complex is impressive.

There's also the pair of ramps to/from Eisenhower Avenue that appear to be unnumbered and considered separately, but they do overlap with the US 1 interchange.

[1995hoo]

....

There's also the pair of ramps to/from Eisenhower Avenue that appear to be unnumbered and considered separately, but they do overlap with the US 1 interchange.

I wasn't counting those because they tie in to a completely separate road (they connect to Mill Road in front of the police station/jail and Mill then connects to Eisenhower, but the latter doesn't connect directly to Route 1). They're also a partial interchange.

[jeffandnicole]

I don't know what rule, inclusion or exclusion this would fall under, but NJ 42 South at CR 534 (https://maps.app.goo.gl/vy9P7bYjmGsN6kvd7, E Church Street/Blackwood-Clementon Rd), stop and yield sign controlled, has a 12 hour traffic count from 2009 of 19,600 vehicles, with 2,400 vehicles at the PM Peak ( https://www.njtms.org/map/tms_reports/reports/9t2-313-wrpt-09_02_26_10.pdf ), and I'm sure traffic hasn't gotten lighter in the 15 years since. 

[Max Rockatansky]

CA 41 at Excelsior Avenue just has stop signs at the end the off ramps.

[BrianP]

I don't know what rule, inclusion or exclusion this would fall under, but NJ 42 South at CR 534 (https://maps.app.goo.gl/vy9P7bYjmGsN6kvd7, E Church Street/Blackwood-Clementon Rd), stop and yield sign controlled, has a 12 hour traffic count from 2009 of 19,600 vehicles, with 2,400 vehicles at the PM Peak ( https://www.njtms.org/map/tms_reports/reports/9t2-313-wrpt-09_02_26_10.pdf ), and I'm sure traffic hasn't gotten lighter in the 15 years since. 

This works without a signal since most people using this interchange are coming from/going to the east which just involve right turns.  People in Blackwood would most likely use the NJ 168 interchange instead. Thus why there's no dedicated left turn lane to the ramp onto 42 north.

[NE2]

I-10 at I-20

[jakeroot]

Personally, I don't think the change makes any sense, because now there is a weave between the two lanes. If the double left had been maintained, there would be no need for a weave since cars could turn left from the lane they needed (either towards 255 [left lane] or Research Drive [right lane]). Traffic entering via the right turn slip lane could have easily turned into either lane depending on where they wanted to go.

I tend to agree with you overall, but it is worth noting that yielding to an unsignalized double-left could be potentially confusing/problematic for those turning right. The flow of left turning traffic would have been fairly jolty and unpredictable, and if it was busy, you'd have to rely on guessing when they would turn and verifying that they were turning into the appropriate lane to accurately judge whether it was safe to proceed.

Point taken. But it operated exactly the way you've described it for over 30 years with no apparent changes to the configuration. That alone doesn't mean it was safe, but I think we can assume it operated with a good safety record if nothing ever changed.

Also, while traffic would have to yield to two oncoming turn lanes, slip lanes almost never have right-of-way. I would argue that yielding to an oncoming double left turn is barely different than yielding to two through lanes from the left. In both cases, you look around as you approach and determine if there is a gap. In both cases, there is a chance that drivers can change lanes, fail to stay in their lane, or whatever. Though with an oncoming double left, you just have much better visibility of the traffic you might have to yield to.

They could have also fixed the issue by just removing the slip lane and making the double left yield to all lanes. In any case, I don't see how the new weave is better or safer; weave locations are notorious danger hot spots.

[TheHighwayMan3561]

I-35 at Central Avenue in Duluth, which abuts a busy commercial district on the north side of the freeway. Also at 21st Ave E, where during college move-in weekend traffic backs up onto the freeway.

In downtown Minneapolis, the ramp from I-35W to 11th Ave ends at a stop sign.

[DandyDan]

Hamilton Street @ US 75 in Omaha on the North Freeway.

[Rothman]

I-35 at Central Avenue in Duluth, which abuts a busy commercial district on the north side of the freeway.

Heh.  "Busy" is quite a relative term.  Super One and the Jade Fountain (heard the latter was having a hard time staying open).

[webny99]

I would argue that yielding to an oncoming double left turn is barely different than yielding to two through lanes from the left. In both cases, you look around as you approach and determine if there is a gap. In both cases, there is a chance that drivers can change lanes, fail to stay in their lane, or whatever. Though with an oncoming double left, you just have much better visibility of the traffic you might have to yield to.

Regarding the bolded phrase, the biggest difference at this location specifically is predictability. Because the left turn is unsignalized, right turning traffic can only guess when left turners might actually complete their turn. Sure, they can *try* to intuit based on the gaps in thru traffic, but it's difficult to keep an eye on traffic behind you from a turning position, and each left turner has a different tolerance for when it's safe to complete the turn, so it's much more of a guessing game than two thru lanes, even despite better visibility.

Also previously, both directions of turning traffic could choose either lane to turn into, since they ended up merging downstream. Now, their lane choice depends on whether they're headed to 255 North or Research Dr. If there's two left turn lanes, now you're asking drivers to pre-select their destination, requiring additional signage, or surprising them into a quick weave after they complete their turn. And it could add delays for right turners, since they now have to make sure both lanes are clear if they're heading to 255 North.

In short: a weave obviously isn't ideal, but I can see how it's the simpler and more effective solution here.

In any case, I don't see how the new weave is better or safer; weave locations are notorious danger hot spots.

Aside from my thoughts above, it seems likely that the responsible DOT identified either a problem with the double left or a benefit to the weave, or they wouldn't have bothered with changing it. Or perhaps more likely, the volumes just didn't warrant two left turn lanes anyways.

[webny99]

I don't know what rule, inclusion or exclusion this would fall under, but NJ 42 South at CR 534 (https://maps.app.goo.gl/vy9P7bYjmGsN6kvd7, E Church Street/Blackwood-Clementon Rd), stop and yield sign controlled, has a 12 hour traffic count from 2009 of 19,600 vehicles, with 2,400 vehicles at the PM Peak ( https://www.njtms.org/map/tms_reports/reports/9t2-313-wrpt-09_02_26_10.pdf ), and I'm sure traffic hasn't gotten lighter in the 15 years since. 

This works without a signal since most people using this interchange are coming from/going to the east which just involve right turns.  People in Blackwood would most likely use the NJ 168 interchange instead. Thus why there's no dedicated left turn lane to the ramp onto 42 north.

Aside from being a partial interchange, this works very similarly to the example in the OP: It's able to sustain high volumes without a traffic signal because the high volume movements are all right turns. Essentially add the missing ramps and it probably ends up pretty similar to the I-490/NY 31 volume and capacity-wise.

[michiganguy123]

I-75 at US-33 Ohio

You have to stop at a stop sign to turn left to get on the US-33 freeway


I-96 at 3 Mile RD/Alpine Ave Michigan
You have to stop at a stop sign and turn left on a 5 lane road to get to the busy Alpine Ave shopping area


As for mainly right turn stops I don't think they should count.

[webny99]

I-75 at US-33 Ohio

You have to stop at a stop sign to turn left to get on the US-33 freeway

I-71 at US 30 near Mansfield, OH is also like this.. but the movements between east and north aren't allowed there and are instead serviced by a separate interchange. This one is a better example because all turn movements are allowed.

I-96 at 3 Mile RD/Alpine Ave Michigan
You have to stop at a stop sign and turn left on a 5 lane road to get to the busy Alpine Ave shopping area

Wow, it is surprising that the left turn here is unsignalized. This is also just a cool interchange design in general, but there is also a stoplight on Alpine Ave in the median of I-96.

As for mainly right turn stops I don't think they should count.

Examples with heavy right turning traffic may be discounted or excluded if you're looking specifically for the *worst* interchange with no stoplights, but that isn't the premise of the thread.

Simply the busier the better is what's being looked for here.

[freebrickproductions]

Plenty of diamonds and other styles of interchange across the rural parts of Alabama only have stop signs for the ramps.

[webny99]

Plenty of diamonds and other styles of interchange across the rural parts of Alabama only have stop signs for the ramps.

Of course, and that's true nation-wide, but how busy they are is the key question here.

EDIT: the quick formula for calculating the approximate AADT of an interchange is:

• Add the surface street counts from either side of the interchange. Divide by 2.
• Add all of the ramp counts.
• Add your answers from 1. and 2.