Improving the Parclo

[tradephoric]

Everyone is use to the Parclo interchange but are there any variations that can make it more efficient?  Here's an example of a Parclo A4 where traffic lights along the arterial only stop one direction of traffic. 

https://www.google.com/maps/@43.97657,-92.4627,318m/data=!3m1!1e3

There are a few real-world examples like this, but could a similar design be applied to an interchange with heavy left turning traffic off the freeway that require 2-lanes of left turning off-ramp traffic?

https://www.youtube.com/watch?v=nCk9iULJ0qE

[NE2]

You can remove all the ramps. Then there will be no delay at the interchange. Just don't expand your study area to include adjacent interchanges.

[kphoger]

Everyone is use to the Parclo interchange but are there any variations that can make it more efficient?  Here's an example of a Parclo A4 where traffic lights along the arterial only stop one direction of traffic. 

https://www.google.com/maps/@43.97657,-92.4627,318m/data=!3m1!1e3

There are a few real-world examples like this, but could a similar design be applied to an interchange with heavy left turning traffic off the freeway that require 2-lanes of left turning off-ramp traffic?

https://www.youtube.com/watch?v=nCk9iULJ0qE

Honestly, I wish all ParClo interchanges had lights that only stopped one direction of travel like that.

[tradephoric]

It is important to consider the adjacent traffic signals in any service interchange design.  Here's a typical Parclo A4 interchange from Toronto.  If the interchange traffic signals stops both directions of travel, it basically ruins any chance of good progression with the adjacent traffic signals (only one direction of the arterial can be timed for smooth flow).  But if the interchange traffic signals stop only one direction of travel, they can be timed for smooth flow along the main arterial in both directions.





EDIT:  One problem with that SYNCHRO model is that traffic exiting the freeway turning left onto the arterial only has so much time to merge to the right most lane of the arterial if they got to make a right turn shortly after the interchange.  But they would still have a good 1200-1500 feet to merge over.

[cbeach40]

Not being a fan of parclos, I want to like this. But while it definitely provides superior operational efficiency to a traditional Parclo on paper (or in Synchro as it were), it's going to take up so much more real estate. That Minnesota example works as it's already a divided highway and the volumes are presumably low (given the single lane for the EB-NB movement). If you take that Ontario example, the NB offramp's left turn volume is about 100 vehicles in the peak hour, so you're definitely going to need that dual left. Factor in the extra space for transport's turning radius and you're going to need to add about 10 m of width to your roadway. Bring that down to 7.5 across the structure (which you'd need to widen). But that's still a lot of extra roadway width.

EDIT:  One problem with that SYNCHRO model is that traffic exiting the freeway turning left onto the arterial only has so much time to merge to the right most lane of the arterial if they got to make a right turn shortly after the interchange.  But they would still have a good 1200-1500 feet to merge over.

If you're designing a five lane move they'll need a good 3000', probably more depending on the arterial speed. Steeles having a high design speed it would need 4100' for that move.


Overall this design would work on low volumes with lots of real estate, but if you have high volumes and intersection density there's definitely some other alternatives that would work better.

[tradephoric]

I put together a SYNCHRO model of the recently constructed DDI in Florida and compared it to the Parclo A4 interchange (w/traffic signals that only stop one direction of traffic).  Each model has the exact same traffic counts (3000 vph along the arterial and 1000 vph for all exiting and entering freeway movements).  The nice thing about the Parclo A4 design is that all traffic entering the freeway from the arterial doesn't stop at traffic signals.  That's huge to keep traffic flowing through the interchange. 

https://www.youtube.com/watch?v=kqfC3wjE6oc

https://www.youtube.com/watch?v=VUx6VwB1gSI

[cbeach40]

I put together a SYNCHRO model of the recently constructed DDI in Florida and compared it to the Parclo A4 interchange (w/traffic signals that only stop one direction of traffic).  Each model has the exact same traffic counts (3000 vph along the arterial and 1000 vph for all exiting and entering freeway movements).  The nice thing about the Parclo A4 design is that all traffic entering the freeway from the arterial doesn't stop at traffic signals.  That's huge to keep traffic flowing through the interchange. 

I'll admit I got a bit confused by its application in that previous example. This seems a lot better, makes more sense.

Just a few questions:
- Where is this interchange located? - edit - Found it.
- Can you you just post a snapshot for the Synchro report files? I'm curious to see how they work head to head.
- What are the left turn volumes on the offramps? I suppose that would be included in the above.
- If you eliminate the force off how does it do then? I'm not a fan of force off (well, I can live with a single but a double raises eyebrows). How would it perform without EB/WB lane 3 being forced onto the loop ramps? - edit - I see from the above link why you had to do that based on the existing structures. For the sake of argument, let's say this is new build, how would it perform without the force off?
- What's the radius of those loop ramps?

Thanks!

[webny99]

One thing about certain parclos that I dislike, is when the non-loop on-ramp traffic has to stop at the traffic signal at the off-ramp. Just extend the ramp, or even an attached lane, so that movement can flow freely. No use making them stop and lose a whole light cycle when no one is going to be conflicting with their movement.

Of course, separating the two flows of traffic means you can't get off the freeway and right back on. But you can pull a u-turn if you really, really have to do that, which should be rare.

[tradephoric]

One thing about certain parclos that I dislike, is when the non-loop on-ramp traffic has to stop at the traffic signal at the off-ramp. Just extend the ramp, or even an attached lane, so that movement can flow freely. No use making them stop and lose a whole light cycle when no one is going to be conflicting with their movement.

Of course, separating the two flows of traffic means you can't get off the freeway and right back on. But you can pull a u-turn if you really, really have to do that, which should be rare.

Here are some real world examples of Parclo A4's with dedicated on-ramp lanes.  Is that what you were picturing?

https://www.google.com/maps?ll=27.41604,-80.38808&z=18&t=h
https://www.google.com/maps/@25.8093231,-80.3215405,235m/data=!3m1!1e3

[webny99]

One thing about certain parclos that I dislike, is when the non-loop on-ramp traffic has to stop at the traffic signal at the off-ramp. Just extend the ramp, or even an attached lane, so that movement can flow freely. No use making them stop and lose a whole light cycle when no one is going to be conflicting with their movement.

Here are some real world examples of Parclo A4's with dedicated on-ramp lanes.  Is that what you were picturing?
https://www.google.com/maps?ll=27.41604,-80.38808&z=18&t=h
https://www.google.com/maps/@25.8093231,-80.3215405,235m/data=!3m1!1e3

Yes, exactly. Traffic using those on-ramps doesn't have to stop at the light at the off-ramp. This is not the case at many parclos in Ontario, and I find it bothersome.

[tradephoric]

- Can you you just post a snapshot for the Synchro report files? I'm curious to see how they work head to head.

I haven't made too many reports in SYNCHRO but here are the simulation summary for the DDI (left side) and Parclo A4 (right side).  I'm not exactly sure why the travel distance and fuel used is higher in the Parclo A4 but i think it has something to do with the fact that the Parclo A4 model has the freeway traffic included, whereas the DDI model didn't.  But looking at total stops and total delay, it's much less in the Parclo model.

- What are the left turn volumes on the offramps? I suppose that would be included in the above.

I searched for the actual Peak Hour volumes for the Florida DDI but was unable to find the environmental assessment online.  Instead i just plugged in 1000 vph volumes for all off-ramp and on-ramp turning movements for both models (which i felt was a high value that would stress each model... and it was easy to input).   

- What's the radius of those loop ramps?

The radius is roughly 250 feet (500 feet diameter).  Probably a little tighter than you would want to see.  One reason i included the freeway traffic in the Parclo model was because i wanted to show how far apart the existing bridge decks are apart.  You could fit two bridge decks in the median of the existing freeway (and as part of the DDI construction both bridge decks were replaced anyways... in fact, as part of the construction they had to temporarily build a bridge deck that was right next to the existing bridge deck).  Had the interchange been designed with the freeway bridges being much closer together, it gives you quite a bit more space for those loop ramps.

[tradephoric]

- If you eliminate the force off how does it do then? I'm not a fan of force off (well, I can live with a single but a double raises eyebrows). How would it perform without EB/WB lane 3 being forced onto the loop ramps? - edit - I see from the above link why you had to do that based on the existing structures. For the sake of argument, let's say this is new build, how would it perform without the force off?

Like this?

https://www.youtube.com/watch?v=NNjgSIXovMg

It still seems to hold its own.  Traffic wishing to enter the freeway from the loop ramp can get delayed if the arterial backs up now, but there are still 4-lanes servicing through traffic along the arterial at the traffic signal.  Compared to the first design, i'm not really seeing much higher delay (and there are only 8 lanes under the bridge as opposed to 10). 

[TheStranger]

One thing about certain parclos that I dislike, is when the non-loop on-ramp traffic has to stop at the traffic signal at the off-ramp. Just extend the ramp, or even an attached lane, so that movement can flow freely. No use making them stop and lose a whole light cycle when no one is going to be conflicting with their movement.

Here are some real world examples of Parclo A4's with dedicated on-ramp lanes.  Is that what you were picturing?
https://www.google.com/maps?ll=27.41604,-80.38808&z=18&t=h
https://www.google.com/maps/@25.8093231,-80.3215405,235m/data=!3m1!1e3

Yes, exactly. Traffic using those on-ramps doesn't have to stop at the light at the off-ramp. This is not the case at many parclos in Ontario, and I find it bothersome.

I THINK California has intentionally set up some new parclos with all traffic heading to onramps having to stop in order to make it more pedestrian-friendly (as opposed to pedestrians having to cross a free-flowing onramp).  Example that comes to mind, a former cloverleaf in San Mateo at the junction of Route 92 and Route 82/El Camino Real:
https://www.google.com/maps/place/Bovet+Rd+%26+S+El+Camino+Real,+San+Mateo,+CA+94402/@37.550207,-122.3124394,18.41z/data=!4m5!3m4!1s0x808f9e5edc9d1a93:0xef334443a88b8463!8m2!3d37.5521644!4d-122.3141689

[mrsman]

One thing about certain parclos that I dislike, is when the non-loop on-ramp traffic has to stop at the traffic signal at the off-ramp. Just extend the ramp, or even an attached lane, so that movement can flow freely. No use making them stop and lose a whole light cycle when no one is going to be conflicting with their movement.

Here are some real world examples of Parclo A4's with dedicated on-ramp lanes.  Is that what you were picturing?
https://www.google.com/maps?ll=27.41604,-80.38808&z=18&t=h
https://www.google.com/maps/@25.8093231,-80.3215405,235m/data=!3m1!1e3

Yes, exactly. Traffic using those on-ramps doesn't have to stop at the light at the off-ramp. This is not the case at many parclos in Ontario, and I find it bothersome.

I THINK California has intentionally set up some new parclos with all traffic heading to onramps having to stop in order to make it more pedestrian-friendly (as opposed to pedestrians having to cross a free-flowing onramp).  Example that comes to mind, a former cloverleaf in San Mateo at the junction of Route 92 and Route 82/El Camino Real:
https://www.google.com/maps/place/Bovet+Rd+%26+S+El+Camino+Real,+San+Mateo,+CA+94402/@37.550207,-122.3124394,18.41z/data=!4m5!3m4!1s0x808f9e5edc9d1a93:0xef334443a88b8463!8m2!3d37.5521644!4d-122.3141689

I think a proper way to address any concerns of pedestrian conflict with ramp traffic would be signals that are limited to the crossing.  The right turn on-ramp should leave the arterial before the traffic signal, so that it does not get slowed down by traffic that it is not in conflict with.

Here is an example of a specialized pedestrian crossing at a channelized right turn.  While this example does not involve an on-ramp or an off-ramp, I could see it being applied in such a case.

https://www.google.com/maps/@33.9958659,-118.3953058,3a,75y,354.85h,79.69t/data=!3m6!1e1!3m4!1smF-RDKK3eAR1BtX1jTZqAg!2e0!7i16384!8i8192

[tradephoric]

Here is an interchange that is a mix between a Parclo AB and a DDI.  The geometry is such that the traffic signals servicing EB traffic entering the freeway should always be in step.  Also, EB Thru traffic doesn't travel through ANY traffic signals.  It seems to be a pretty efficient interchange.

https://www.youtube.com/watch?v=trfc_7cGqJA

The design was inspired from the interchange they are building at I-94 & Sheyenne Street in West Fargo.

https://www.youtube.com/watch?v=yqiJo25h-ak

[jakeroot]

Here is an interchange that is a mix between a Parclo AB and a DDI.  The geometry is such that the traffic signals servicing EB traffic entering the freeway should always be in step.  Also, EB Thru traffic doesn't travel through ANY traffic signals.  It seems to be a pretty efficient interchange.

https://www.youtube.com/watch?v=trfc_7cGqJA

Why no right turn for the "southbound" off-ramp (top-left)?

Can you see if you can compact the interchange size down a bit? Preferably, narrow the arterial a bit. The "eastbound" carriageway could be brought much closer to the "westbound" carriageway without losing functionality.

Here is an example of a specialized pedestrian crossing at a channelized right turn.  While this example does not involve an on-ramp or an off-ramp, I could see it being applied in such a case.

https://www.google.com/maps/@33.9958659,-118.3953058,3a,75y,354.85h,79.69t/data=!3m6!1e1!3m4!1smF-RDKK3eAR1BtX1jTZqAg!2e0!7i16384!8i8192

Odd that they never bothered to add a stop line when the switched the crossing style from transverse to zebra.

[tradephoric]

Why no right turn for the "southbound" off-ramp (top-left)?

Can you see if you can compact the interchange size down a bit? Preferably, narrow the arterial a bit. The "eastbound" carriageway could be brought much closer to the "westbound" carriageway without losing functionality.

Hey Jake, I just forgot to add that SB RT movement in that original model .  Here's a "compact" version of the interchange with that movement added.

https://www.youtube.com/watch?v=FMm8c_EIr98

[jakeroot]

Hey Jake, I just forgot to add that SB RT movement in that original model .  Here's a "compact" version of the interchange with that movement added.

Ahh, no problem. I appreciate the changes. More realistic sizing IMO.

Does the simulation allow for double right turns on red? I see only the right lane turning on red. Don't see any turning on red for the loop off-ramp, actually.

Would like to see those on-ramps modified to include ramp meters, too (though I understand those are still taboo in some areas of the country). Those always fuck with arterials, unless the DOT actually watches them, to prevent such occurrences. I think arterials could be built better to accommodate ramp meter stacking.

[tradephoric]

Does the simulation allow for double right turns on red? I see only the right lane turning on red. Don't see any turning on red for the loop off-ramp, actually.

I'm honestly not sure why the exiting loop traffic isn't turning right on red as the simulation is set to allow right turn on red.  Even when i reduced the arterial volumes traffic still wasn't turning right on red in gaps.  OTOH, when i reduced the arterial traffic volumes i did observe double right turns on red at the non-loop off ramp.

[tradephoric]

The other question i have is do Parclo loop ramps really need to have such large diameters?  There are plenty of parclo interchanges throughout LA with small loop ramp diameters (less than 400 feet and in some cases less than 300 feet).  Here are some examples of LA Parclo's with small diameters off the I-5, I-405, and I-110 freeways.  The first link is interesting as all traffic entering the freeway can bypass all interchange traffic signals:

https://www.google.com/maps/@33.6891819,-117.9191526,368m/data=!3m1!1e3
https://www.google.com/maps/@33.9039749,-118.1857329,368m/data=!3m1!1e3
https://www.google.com/maps?ll=33.95784,-118.11178&z=18&t=h
https://www.google.com/maps?ll=33.85088,-117.81476&z=18&t=h

[kphoger]

The other question i have is do Parclo loop ramps really need to have such large diameters? 

In my opinion, that depends on three things:
1.  Grade
2.  Traffic speed
3.  Commercial traffic

A good example of a tight loop ramp not being sufficient is at Kellogg & I-235 here in Wichita.  The ramp is signed at 15 mph, yet traffic then has to merge onto a 65 mph highway from there.  It's quite a steep grade, and the area sees trucks pulling double and triple trailers.  I've even heard of trucks getting stuck on the ramp in the past.

[tradephoric]

Here is a single-point Parclo B4 interchange.  One operational advantage to this interchange over a typical Parclo B4 interchange is that arterial traffic turning left onto the freeway doesn't get delayed by a traffic signal servicing the opposing left-turning on-ramp traffic.

https://www.youtube.com/watch?v=6vef4O6Vs_A

[kphoger]

Here is a single-point Parclo B4 interchange.  One operational advantage to this interchange over a typical Parclo B4 interchange is that arterial traffic turning left onto the freeway doesn't get delayed by a traffic signal servicing the opposing left-turning on-ramp traffic.

https://www.youtube.com/watch?v=6vef4O6Vs_A

So that's an improvement on your "Folded Interchange", I assume?

[tradephoric]

^I was wondering that question myself. 

From a cost perspective the "folded interchange" would require a wider bridge deck to fit the contraflow lefts but the bridge deck at a Single-Point interchange is often quite wide anyways.  Maybe it's a wash?  Could argue the Single-Point would have less ROW costs depending on how all the on/off ramps are designed.

From an operational advantage, you could argue that loop traffic exiting the freeway would drive a slightly shorter distance at the Single-Point Parclo B4 than they would at the "folded interchange" so delay for that turning movement might be slightly smaller (the same argument could be made for arterial traffic turning left onto the freeway).

From a safety perspective, you definitely wouldn't want a driver to accidentally enter the freeway from an off-ramp.  You could make the case that this dangerous scenario would be more likely to occur at a Single-Point Parclo B4 interchange (but I'm sure there is a design that could mitigate the risk). 

[NoGoodNamesAvailable]

https://www.google.com/maps/@33.9958659,-118.3953058,3a,75y,354.85h,79.69t/data=!3m6!1e1!3m4!1smF-RDKK3eAR1BtX1jTZqAg!2e0!7i16384!8i8192

Odd that they never bothered to add a stop line when the switched the crossing style from transverse to zebra.

Yeah, this is not an intuitive setup at all, and I could imagine this crosswalk having a pretty bad driver intrusion problem if not a red light running problem. I've never liked how western states don't stripe stop bars in advance of standard/ladder crosswalks. It's such a clear operational safety deficiency for very little cost benefit.