How do highway engineers choose between over/underpasses for interchanges?

[kernals12]

Intuitively it would seem like overpasses would be cheaper because it negates the need for excavation, but here in Maricopa County's East Valley, freeways almost always pass under cross streets. Did Mesa, Gilbert, Chandler, and Tempe successfully extort ADOT into paying more to build freeways just for aesthetic benefits or am I wrong and having the highway go under cross streets is cheaper?

[pderocco]

There's a commonly used construction technique where the overpass is built on the ground, and then undercut to make room for the freeway. Meanwhile, the future freeway traffic uses what eventually become the diamond ramps.

You can see a lot of this in the Google Earth historic imagery for the Las Vegas beltway.

[kernals12]

There's a commonly used construction technique where the overpass is built on the ground, and then undercut to make room for the freeway. Meanwhile, the future freeway traffic uses what eventually become the diamond ramps.

You can see a lot of this in the Google Earth historic imagery for the Las Vegas beltway.

And they're doing that now here in AZ on SR 24.

[Beltway]

Intuitively it would seem like overpasses would be cheaper because it negates the need for excavation, but here in Maricopa County's East Valley, freeways almost always pass under cross streets. Did Mesa, Gilbert, Chandler, and Tempe successfully extort ADOT into paying more to build freeways just for aesthetic benefits or am I wrong and having the highway go under cross streets is cheaper?

Mainly topography. The simplest is to have the local road overpass the freeway.

When the topography makes that difficult, that is when a better design is to go under the freeway.

Regarding urban freeways with lots of closely spaced local road crossings, they can choose between passing over all the streets or a depressed highway passing under the streets.

[pderocco]

Intuitively it would seem like overpasses would be cheaper because it negates the need for excavation, but here in Maricopa County's East Valley, freeways almost always pass under cross streets. Did Mesa, Gilbert, Chandler, and Tempe successfully extort ADOT into paying more to build freeways just for aesthetic benefits or am I wrong and having the highway go under cross streets is cheaper?

Mainly topography. The simplest is to have the local road overpass the freeway.

When the topography makes that difficult, that is when a better design is to go under the freeway.

Regarding urban freeways with lots of closely spaced local road crossings, they can choose between passing over all the streets or a depressed highway passing under the streets.

Since most of the Las Vegas beltway, or at least the newer parts, are below grade, I think it may also have to do with sound control. Having the road at a lower level than the sound walls creates a sound shadow that projects upward even more, so the adjacent homes hear it even less, at least for the high frequencies.

[kernals12]

Regarding urban freeways with lots of closely spaced local road crossings, they can choose between passing over all the streets or a depressed highway passing under the streets.

I see. And I suppose that in that scenario, bringing in a whole lot of dirt for embankments is just as expensive as removing dirt for a depressed roadway?

[vdeane]

There's a commonly used construction technique where the overpass is built on the ground, and then undercut to make room for the freeway. Meanwhile, the future freeway traffic uses what eventually become the diamond ramps.

You can see a lot of this in the Google Earth historic imagery for the Las Vegas beltway.

How does that work when the road isn't initially constructed as an overpass?  They'd still have to build the bridge; they can't just excavate under asphalt.

[sprjus4]

There's a commonly used construction technique where the overpass is built on the ground, and then undercut to make room for the freeway. Meanwhile, the future freeway traffic uses what eventually become the diamond ramps.

You can see a lot of this in the Google Earth historic imagery for the Las Vegas beltway.

How does that work when the road isn't initially constructed as an overpass?  They'd still have to build the bridge; they can't just excavate under asphalt.

Construct an overpass with very little clearance directly next to the existing road, shift traffic onto the new bridge, demolish the existing road, then excavate out underneath the new overpass and old road to create the proper 16 ft clearance needed.

[Beltway]

Regarding urban freeways with lots of closely spaced local road crossings, they can choose between passing over all the streets or a depressed highway passing under the streets.

I see. And I suppose that in that scenario, bringing in a whole lot of dirt for embankments is just as expensive as removing dirt for a depressed roadway?

Generally in the same realm for earthwork. The depressed highway would require drainage outfalls to drain rainwaters away, and possibly electric pumps as well.

The elevated highway could also be built on a long bridge that passes over the local streets.

Since most of the Las Vegas beltway, or at least the newer parts, are below grade, I think it may also have to do with sound control. Having the road at a lower level than the sound walls creates a sound shadow that projects upward even more, so the adjacent homes hear it even less, at least for the high frequencies.

Depressed highways definitely are less noisy to the adjacent development. One of the commonly cited benefits.

[kphoger]

Someone really needs to use this thread title as the daily question in the "worst possible answer" thread...

[fillup420]

my initial thought is that its cheaper to overpass the freeway, since that usually means just one bridge instead of two, and then the freeway is on the ground level, meaning any bridge issues would close the intersecting road instead of the freeway itself.

[kernals12]

The elevated highway could also be built on a long bridge that passes over the local streets.

That's way more expensive

[Beltway]

The elevated highway could also be built on a long bridge that passes over the local streets.

That's way more expensive

Than building it on embankment fill? Not necessarily. That still requires mainline bridges over each local road. All that soil fill material has to be excavated from borrow pits, and borrow excavation is typically twice the cost per cubic yard as compared to regular excavation (from the right-of-way).

[pderocco]

There's a commonly used construction technique where the overpass is built on the ground, and then undercut to make room for the freeway. Meanwhile, the future freeway traffic uses what eventually become the diamond ramps.

You can see a lot of this in the Google Earth historic imagery for the Las Vegas beltway.

How does that work when the road isn't initially constructed as an overpass?  They'd still have to build the bridge; they can't just excavate under asphalt.

Yes, they do build a bridge, but on or close to ground level. Perhaps it's just easier to do that than to build it 20 feet in the air. I don't know why, but I've seen it done in several places, including when CA-210 was extended from I-15 to I-215, and I'd occasionally drive along next to the construction zone to see their progress.

[kernals12]

okay, I emailed the Arizona Department of Transportation and got this answer
"depressed profiles for traffic interchanges are generally cheaper since they require less in the way of bridge structures.  In addition, they also tend to be safer since it is easier to maintain horizontal line of sight (stopping distances) and easier to respond to and clear crashes than doing so from bridges or overpasses."

[pderocco]

okay, I emailed the Arizona Department of Transportation and got this answer
"depressed profiles for traffic interchanges are generally cheaper since they require less in the way of bridge structures.  In addition, they also tend to be safer since it is easier to maintain horizontal line of sight (stopping distances) and easier to respond to and clear crashes than doing so from bridges or overpasses."

That sounds like they used AI for that answer.

[kernals12]

Another benefit: when highways are below grade, on-ramps assist the acceleration of cars via gravity and off-ramps assist braking in the same way. It's a form of regenerative braking. That saves fuel and, importantly for trucks, reduces brake wear. When the highway is above grade, it's the opposite, gravity means traffic uses more energy to accelerate onto the highway and brakes harder when getting off.

[D3r3k23]

It looks like the Miami area almost exclusively uses overpasses for freeways instead of underpasses. I'm guessing that's because there are lots of waterways that need bridges anyways.

[Rothman]

It looks like the Miami area almost exclusively uses overpasses for freeways instead of underpasses. I'm guessing that's because there are lots of waterways that need bridges anyways.

It's also flat and I'd also wonder about drainage and the water table.

[Quillz]

One thing I always found interesting in the SF Valley is 101 is almost exclusively overpassing the streets, while 118 is almost exclusively underpassing the streets.

The only exception on 101 is Exit 29, Valley Circle/Mulholland. Here, 101 underpasses the street, but everything else all the way to the Hollywood Freeway is an overpass, with the exception of some streets that do not have exits.

118 is basically the inverse. At the western edge of the valley, Topanga and DeSoto underpass the freeway, but everything else (again, minus a few streets without exits) tend to overpass the freeway. If you look at the streets northward from the 101, you can clearly see the elevation gradually rises (as the streets appear to dip a little but then rise). I'm guessing the slightly higher geography is why the 118 tends to run under the streets, but the freeway is also quite a bit newer than the 101, not being fully completed until 1994 if I remember correctly.

[D-Dey65]

Intuitively it would seem like overpasses would be cheaper because it negates the need for excavation, but here in Maricopa County's East Valley, freeways almost always pass under cross streets. Did Mesa, Gilbert, Chandler, and Tempe successfully extort ADOT into paying more to build freeways just for aesthetic benefits or am I wrong and having the highway go under cross streets is cheaper?

Mainly topography. The simplest is to have the local road overpass the freeway.

When the topography makes that difficult, that is when a better design is to go under the freeway.

Regarding urban freeways with lots of closely spaced local road crossings, they can choose between passing over all the streets or a depressed highway passing under the streets.

I've always looked at it that way. Sometimes I can just look at the land around an intersection with a highway, and know right away which road should go over or under. That's how I know that Horse Block Place in Farmingville, New York near the Suffolk County Community College should go over Suffolk CR 97 instead of under it, with no access.

[architect77]

This reminds me of the lower cost alternative to stack interchanges  that supposedly cost billions now for multiple flyovers.

With all of central NC growing and more interchanges planned for the future,  NCDOT will limit the number of new flyovers to existing locations and all new interchanges will instead feature less impressive -looking turbine interchanges that stay on the ground as much as possible.

Imagine how cool this 500 acre turbine interchange in Raleigh/Wake Co. would be as a stacked structure.

You don't even experience this complex junction as such because the ramps are always on the ground and intersecting roadways are just dug underneath.


540turbine by Stephen Edwards, on Flickr

[kphoger]

the ramps are always on the ground

that's gotta suck

[sprjus4]

This reminds me of the lower cost alternative to stack interchanges  that supposedly cost billions now for multiple flyovers.

With all of central NC growing and more interchanges planned for the future,  NCDOT will limit the number of new flyovers to existing locations and all new interchanges will instead feature less impressive -looking turbine interchanges that stay on the ground as much as possible.

Imagine how cool this 500 acre turbine interchange in Raleigh/Wake Co. would be as a stacked structure.

You don't even experience this complex junction as such because the ramps are always on the ground and intersecting roadways are just dug underneath.


540turbine by Stephen Edwards, on Flickr

I personally think that interchange is awesome. Same with the I-485 / I-85 turbine.

[architect77]

This reminds me of the lower cost alternative to stack interchanges  that supposedly cost billions now for multiple flyovers.

With all of central NC growing and more interchanges planned for the future,  NCDOT will limit the number of new flyovers to existing locations and all new interchanges will instead feature less impressive -looking turbine interchanges that stay on the ground as much as possible.

Imagine how cool this 500 acre turbine interchange in Raleigh/Wake Co. would be as a stacked structure.

You don't even experience this complex junction as such because the ramps are always on the ground and intersecting roadways are just dug underneath.


540turbine by Stephen Edwards, on Flickr

I personally think that interchange is awesome. Same with the I-485 / I-85 turbine.

i don't dislike these turbine designs, it just would look impressive if it were like a 5 level stack. I lived in LA with many multi level stacks and I've driven through many in Texas, but Atlanta's spaghetti junction is my favorite because it has a lot of symmetry. The upcoming express lanes will tie into it with 4-5 additional flyovers above the existing elevated ramps.

I don't know how the environmental protectionists feel about so much land used for this particular interchange.

Increased use of grade separation seems helpful for overused arterials too. Briefly rising over cross streets that used to be managed with traffic lights taking 2-3 minutes to cycle through seems to be a significant although expensive improvement effort.