Introducing the Sparclo, an SPUI/Parclo hybrid (with mocked up US example)

[jackal]

Here's an idea for a very high capacity interchange between a motorway and surface street, which I call a sparclo - a portmanteau of SPUI and parclo. First, a sketch:

Like some six-ramp parclos (6RP), six of eight turns are freeflow (e.g., Floridian 6RP). However, the sparclo improves on the 6RP by allowing the remaining two long turns (left in the US) to proceed simultaneously, passing side-by-side in SPUI fashion. While the 6RP requires two intersections, the sparclo only requires one, reducing journey times for both long turns and through traffic on the surface street compared to a 6RP.

We can also compare the sparclo to the SPUI. In this comparison the sparclo's key advantage is that two of the long turns are freeflow. This is not only good for these long turns, but good for the other long turns and surface street through traffic, as the signalised junction now only requires two phases (one for through traffic, another for the pair of long turns) rather than the three required in an SPUI (one for through traffic, another for the first pair of long turns, a third for the other pair of long turns). Signage and markings are also simplified as you only have four paths of traffic using the central intersection rather than six.

The 6RP and SPUI are the highest capacity non-system interchanges currently used, and the sparclo clearly has significantly higher capacity than either. It still only requires one bridge as these junctions do and should be significantly cheaper than system interchanges.

I drew the sketch above as a modification of the SPUI at I-269/SR 57 in Piperton, TN. I chose this site as it has space for the loops unlike most SPUIs, as well as extra bridge spans to utilise (though a quirk of the site is that a bridge is required to cross the railway). I don't think there's really any need for such an upgrade there - it's just to show how it might be retrofitted to an SPUI where the space for the loops is available. New builds are also possible of course.

[Bryn666]

This looks absolutely bonkers, yet... I suspect it would work rather well.

Given the alignment of the slip roads as well, you could fit in Dutch style underpasses and in suburban locations (I'm looking at you, M8), there'd be minimal severance for active travel either.

[Vierwielen]

On the plus side, it will replace a four-level intersection with a three level intersection, but without any roundabouts (which are the bane of many British motorway intersections.

On the negative side, it will take up for more land, particularly as a full circles need to be accomodated within the roundabout.

I don't know about maintenance costs, but I suspect that the Sparco woudl be cheaper to build, but the land would cost more. It is a trade -off.

[jackal]

This looks absolutely bonkers, yet... I suspect it would work rather well.

Given the alignment of the slip roads as well, you could fit in Dutch style underpasses and in suburban locations (I'm looking at you, M8), there'd be minimal severance for active travel either.

Indeed. Also some of the short turns (right in the US) could be signalised for NMU purposes, as at many Canadian parclos. I drew the highest capacity version, but it's a solid upgrade on a parclo or SPUI even with some extra signals.

On the plus side, it will replace a four-level intersection with a three level intersection, but without any roundabouts (which are the bane of many British motorway intersections.

On the negative side, it will take up for more land, particularly as a full circles need to be accomodated within the roundabout.

I don't know about maintenance costs, but I suspect that the Sparco woudl be cheaper to build, but the land would cost more. It is a trade -off.

It is a two-level intersection, the same as a parclo or SPUI. In the drawing the existing SPUI is removed and the new stuff is on that level - nothing's built above it.

There's no roundabout - my lazy way of drawing a constant radius loop in MS Paint is to draw a circle then grey out the bit that shouldn't be there.

Land take would be a lot higher than an SPUI, mostly due to the loops, and slightly higher than a 6 ramp parclo.

I've maybe drawn and described it in an over-complicated way. It's basically just a parclo with two unhooked right turns - which is actually a pretty big upgrade.

[nowster]

https://maps.app.goo.gl/CA9D2i9DQgbCi7W3A M50 J5 (IRL) comes close.

[Bryn666]

This looks absolutely bonkers, yet... I suspect it would work rather well.

Given the alignment of the slip roads as well, you could fit in Dutch style underpasses and in suburban locations (I'm looking at you, M8), there'd be minimal severance for active travel either.

Indeed. Also some of the short turns (right in the US) could be signalised for NMU purposes, as at many Canadian parclos. I drew the highest capacity version, but it's a solid upgrade on a parclo or SPUI even with some extra signals.

On the plus side, it will replace a four-level intersection with a three level intersection, but without any roundabouts (which are the bane of many British motorway intersections.

On the negative side, it will take up for more land, particularly as a full circles need to be accomodated within the roundabout.

I don't know about maintenance costs, but I suspect that the Sparco woudl be cheaper to build, but the land would cost more. It is a trade -off.

It is a two-level intersection, the same as a parclo or SPUI. In the drawing the existing SPUI is removed and the new stuff is on that level - nothing's built above it.

There's no roundabout - my lazy way of drawing a constant radius loop in MS Paint is to draw a circle then grey out the bit that shouldn't be there.

Land take would be a lot higher than an SPUI, mostly due to the loops, and slightly higher than a 6 ramp parclo.

I've maybe drawn and described it in an over-complicated way. It's basically just a parclo with two unhooked right turns - which is actually a pretty big upgrade.

Definitely can see merit in this, I may have to do you a more detailed drawing.

[jackal]

https://maps.app.goo.gl/CA9D2i9DQgbCi7W3A M50 J5 (IRL) comes close.

This is an offside cloverstack.

Definitely can see merit in this, I may have to do you a more detailed drawing.

Would be interesting to see. Maybe best to do it the other way round from what I had, i.e., loops diverge from the surface street rather than motorway. Ontario at least think it's better to stop traffic off the freeway at signals.

[Peter Freeman]

I drew the sketch above as a modification of the SPUI at I-269/SR 57 in Piperton, TN. I chose this site as it has space for the loops unlike most SPUIs, as well as extra bridge spans to utilise (though a quirk of the site is that a bridge is required to cross the railway). I don't think there's really any need for such an upgrade there - it's just to show how it might be retrofitted to an SPUI where the space for the loops is available. ...

Oh, that one's a SPUI onto a single cariageway cross-street. Quite a light load, so it's unusual. It has space to widen the cross-street, even without the parclo conversion.

[ChrisH]

I like the concept.

You could save some land-take in the SW and NE quadrants by tightening the freeflow right turns off the surface street and aligning them to the signalised left turns.

I also wonder how much more efficient in signal staging it would be compared to that Florida example, where presumably the two left turns are green at the same time and the central area is a reservoir holding the traffic.

[jackal]

You could save some land-take in the SW and NE quadrants by tightening the freeflow right turns off the surface street and aligning them to the signalised left turns.

Good spot. I drew it like that so I could put the sliproads off the loops through the spare spans (e.g., to the right of the support here: https://www.google.co.uk/maps/@35.04440 ... ?entry=ttu). But yes, usually it can be tighter.

[Peter Freeman]

But yes, usually it can be tighter.

And in many existing SPUI locations, it would need to be tighter.

[Peter Freeman]

Your Sparclo is a good idea Jackal, especially for a new-build.

Regarding conversion of existing SPUI's, they tend to be located where there simply isn't room for loops, which is one reason why they're popular and appropriate. In which case, Parclo-ization won't fit.

I also wonder how much more efficient in signal staging it would be compared to that Florida example, where presumably the two left turns are green at the same time and the central area is a reservoir holding the traffic.

Yes Chris, SPUI-ization of an existing 6-ramp parclo might not produce great gains because, at most of them, the hooked long turns are lightly loaded. Light loading of the two at-grade long turns especially occurs where two main roads cross at a skew, which is one of the best reasons for selecting a 6-ramp parclo (as at Tonkin Hwy, Perth WA, Jackal). With light loading, the usual problem (a green phase runs out of stored vehicles, as at a standard diamond service interchange) doesn't occur. And, as you surmise, Chris, the two at-grade long-turn greens can occur simultaneously, so there can be 2-stage operation, and so single-point provides little gain.

Having said all that, the Sparclo's still a good idea!

[Peter Freeman]

... Maybe best to do it the other way round from what I had, i.e., loops diverge from the surface street rather than motorway. Ontario at least think it's better to stop traffic off the freeway at signals.

I've wondered quite a lot about that question of which way around is best. I'm still undecided. Local factors, including space availability, might usually dictate.

[jackal]

Thanks Peter. I agree, the vast majority of SPUIs will not have space for a sparclo. Still, it's an upgrade option in unusual cases of SPUIs with space. New build or full reconstruction are the main use cases.

I also wonder how much more efficient in signal staging it would be compared to that Florida example, where presumably the two left turns are green at the same time and the central area is a reservoir holding the traffic.

Yes Chris, SPUI-ization of an existing 6-ramp parclo might not produce great gains because, at most of them, the hooked long turns are lightly loaded. Light loading of the two at-grade long turns especially occurs where two main roads cross at a skew, which is one of the best reasons for selecting a 6-ramp parclo (as at Tonkin Hwy, Perth WA, Jackal). With light loading, the usual problem (a green phase runs out of stored vehicles, as at a standard diamond service interchange) doesn't occur. And, as you surmise, Chris, the two at-grade long-turn greens can occur simultaneously, so there can be 2-stage operation, and so single-point provides little gain.

One point I would mention is that the "reservoir" isn't only filling up from long turns, but also from straight-through traffic on the surface street. So even if the two non-loop long turns are relatively light, it could be that the reservoir capacity is exceeded if the street has high volumes. The sparclo would be helpful in such cases.

Additionally, it's actually quite unusual for junctions to neatly have the two busy long turns in opposite corners. Yes, sometimes this is the case at skewed junctions, but even then often not, and most 6RPs are not skewed in any case. For more balanced flows, or where you have three busy long turns, or two busy adjacent long turns, and so on, there can be significant gains from the 6RP.

Finally, it's relevant that the cost and land take of the sparclo are not much different from the 6RP. For a new build, even if there isn't much benefit, or the benefit would only be felt a couple of decades later when the reservoir starts to fill, it could be a good option.

[jackal]

Here's a broader thought. The sparclo can be considered as doing much the same as a stacked SPUI, but prioritizing turns rather than straight-ahead movements. That is, if we run through the twelve movements, they provide them in the same basic way:

8 freeflow
4 at SPUI-like signals

The differences only become apparent when we get more detailed. A stacked SPUI does this:

8 freeflow (4 short turns, 4 straight-aheads)
4 at SPUI-like signals (4 long turns)

A sparclo does this:

8 freeflow (4 short turns, 2 straight-aheads, 2 long turns)
4 at SPUI-like signals (2 straight-aheads, 2 long turns)

Given the stacked SPUI is a pretty good design, and the sparclo is providing a broadly similar scale of provision while only using two level rather than three, I think the sparclo has quite a lot going for it. Of course, they are suited to different circumstances, and the SPUI does get something for the extra level - two of the freeflow movements are on much better mainline alignments than the sparclo's looped equivalents.

It's interesting, to me at least, that seemingly completely different designs have quite a lot in common when you look under the bonnet.

[Peter Freeman]

I also wonder how much more efficient in signal staging it would be compared to that Florida example, where presumably the two left turns are green at the same time and the central area is a reservoir holding the traffic.

Yes Chris, SPUI-ization of an existing 6-ramp parclo might not produce great gains because, at most of them, the hooked long turns are lightly loaded. Light loading of the two at-grade long turns especially occurs where two main roads cross at a skew, which is one of the best reasons for selecting a 6-ramp parclo (as at Tonkin Hwy, Perth WA, Jackal). With light loading, the usual problem (a green phase runs out of stored vehicles, as at a standard diamond service interchange) doesn't occur. And, as you surmise, Chris, the two at-grade long-turn greens can occur simultaneously, so there can be 2-stage operation, and so single-point provides little gain.

One point I would mention is that the "reservoir" isn't only filling up from long turns, but also from straight-through traffic on the surface street. So even if the two non-loop long turns are relatively light, it could be that the reservoir capacity is exceeded if the street has high volumes. The sparclo would be helpful in such cases.

"...but also from straight-through traffic on the surface street ..."
I'm not sure that I understand your argument here.

Let's assume (in order to keep the discussion simple) that the standard 6RP is in the UK and is the way around where a N-S motorway feeds the loops via its exit ramps. These free-flow long turns never stop, so neither of them is filling up a reservoir. The reservoir consists of only the right-hand lanes of the cross street between the two at-grade intersections. Straight-through E-W and W-E vehicles must stop occasionally for turning traffic to cross right-to-left at-grade, but not for long, and they do not feed the reservoir: they subsequently drive past red right arrows. The only source of vehicles into the reservoirs is E-W and W-E traffic wanting to turn right. Typically, neither of these movements will be high flow, as appropriate design will cause loops to serve the high flows.

This is why I think the Sparclo is a good idea, but, in most cases, will produce only a moderate capacity increase.

[Peter Freeman]

Here's a broader thought. The sparclo can be considered as doing much the same as a stacked SPUI, but prioritizing turns rather than straight-ahead movements. ...

The observations and comparisons in your analysis are indeed fascinating. The stacked SPUI and the Sparclo (or even standard 6RP) are worthy of much more usage. BTW, Perth's Tonkin Hwy will get another 6RP in its next southern extension, at 2m25s into this video -
https://www.youtube.com/watch?v=UJivUiUrpA8.

... It's interesting, to me at least, that seemingly completely different designs have quite a lot in common when you look under the bonnet.

Interesting to me too. A bit like 'morphing' one type to another merely by pulling a connector into a different quadrant.

[Peter Freeman]

I've been thinking about the space requirement for the single point in your new interchange. In a conversion from a SPUI, it will fit (in the same place as the SPUI's), but in a conversion from a 6RP it will be tight to get the angles and width right (in your example with the spare spans you got lucky). Even in a new build, it will require a more spacious bridge than a standard 6RP.

A solution occurs to me though: the single point crossover doesn't actually need to be on (or under) the bridge. It could be almost anywhere along the cross-street by extending one of the ramps in parallel. The crossover, and the tails of the loops, would bulge out in otherwise unused space. Then, the bridge itself need be wider only to the degree of carrying the ramp extension. I could explain this better, if necessary.

[Peter Freeman]

Definitely can see merit in this, I may have to do you a more detailed drawing.

A proper drawing would be useful, Bryn. I have concerns that it will be hard to actually implement in real-life.

Your drafting of the M67 SPUI idea was really good, and was useful as a verification of feasibility.

[Bryn666]

Definitely can see merit in this, I may have to do you a more detailed drawing.

A proper drawing would be useful, Bryn. I have concerns that it will be hard to actually implement in real-life.

Your drafting of the M67 SPUI idea was really good, and was useful as a verification of feasibility.

We need a UK location that would be suitable for this challenge I think.