New and Expanded Highways

[JeffDowntown]

I've never driven through a diverging diamond interchange but it does appear as if traffic could merge more smoothly although lights are now involved. The current cloverleaf interchanges are just terrible so almost anything would be an improvement.

It looks like Providence added one recently;
https://www.google.com/maps/place/W...4c9703afb77663!8m2!3d41.7001009!4d-71.4161671

The diverging diamond interchanges work really well. There is a relatively new one on I 40 near Winston Salem North Carolina. It solved a lot of former exiting traffic backup issues onto the highway travel lanes. Cars clear through the interchange much faster!

http://www.journalnow.com/news/loca...cle_a1887414-d580-5c33-8b4a-a05898706136.html

 

[Scalziand]

SPUIs are pretty nice too.

 

[Arlington]

SPUIs are pretty nice too.

Isn't a diverging diamond just a pair of SPUIs?

At any rate, yes, land tied up in the loops of cloverleafs is all wrong:
- too much extremely valuable land
- too many conflicting movements

Really bad mobility-per-acre.

 

[omaja]

Isn't a diverging diamond just a pair of SPUIs?

No, they two serve different purposes. In essence, a SPUI is a better all-around interchange whereas the DD is better for instances where traffic is more unidirectional at any given time (say morning versus evening rush). SPUIs centralize all movements in one very high-capacity intersection with three light phases which allows all movements equal priority. A DD just eliminates the left-turn phase of a normal diamond interchange, prioritizing traffic traveling on the surface road and entering the freeway (at the expense of traffic exiting which yields to the surface road on both left and right movements).

 

[Joel N. Weber II]

Single Point Urban Interchanges vs Diverging Diamond Interchanges

I'm failing to find any really clear references for this, but I thought I'd seen things that suggest that SPUIs tend to be prone to crashes when drivers get confused by the design, whereas DDIs seem to be safer. Also, I think SPUIs may end up requiring even more width on the secondary road than a conventional diamond interchange (IIRC there's a DDI in Connecticut visible on the satellite view in Google Maps that seems to make this clear), whereas in some cases a diverging diamond interchange can work with a narrower road than a conventional diamond interchange, so my understanding is that DDIs tend to be a win if keeping bridge costs down is a goal.

 

[Scalziand]

I'm not aware of any DDI's in CT. Are you referring to the SPUI in Trumbull? CTDOT did have to significantly increase the width of the Rt111 bridge over the Merritt Parkway in Trumbull when they converted it from a regular diamond to a SPUI.

 

[Joel N. Weber II]

Yes, sorry, where I typed ``DDI in Connecticut'' I had intended to type ``SPUI in Connecticut'' referring to the one in Trumbull.

 

[Joel N. Weber II]

US 1 in Newburyport / across the Merrimack

Is there any traffic count data available that would indicate whether one automobile travel lane in each direction on the US 1 bridge that crosses the Merrimack would be sufficient? (Notice that US 1 just north of the bridge in Salisbury does narrow to one lane in each direction, although some bridge traffic may take 1st / Ferry instead.)

I'm wondering whether the conversion of the right lane from a general purpose travel lane to a bicycle lane, as happened / will happen after construction finishes on the Longfellow bridge in the Boston to Cambridge direction, would be viable on that bridge across the Merrimack.

(Also, is the current surface on that Merrimack River bridge bike friendly?)

Does the US 1 bridge over Merrimac St in Newburyport see traffic volumes that make it more valuable than the Casey Overpass?

 

[The EGE]

Re: US 1 in Newburyport / across the Merrimack

Is there any traffic count data available that would indicate whether one automobile travel lane in each direction on the US 1 bridge that crosses the Merrimack would be sufficient? (Notice that US 1 just north of the bridge in Salisbury does narrow to one lane in each direction, although some bridge traffic may take 1st / Ferry instead.)

http://lmgtfy.com/?q=massdot+traffic+counts

 

[tangent]

I've never driven through a diverging diamond interchange but it does appear as if traffic could merge more smoothly although lights are now involved. The current cloverleaf interchanges are just terrible so almost anything would be an improvement.

It looks like Providence added one recently;
https://www.google.com/maps/place/W...4c9703afb77663!8m2!3d41.7001009!4d-71.4161671

Cloverleaf interchanges work well enough for quite a few exits. But there is a big difference between the 95/93 cloverleaf and the 95/rt 4 cloverleaf and the like. Crazy that 95/93 interchange has a pattern designed for far fewer cars.

 

[omaja]

I'm failing to find any really clear references for this, but I thought I'd seen things that suggest that SPUIs tend to be prone to crashes when drivers get confused by the design, whereas DDIs seem to be safer.

I've read SPUIs are just as safe as a traditional diamond interchange but are built to handle significantly more volume. Perhaps this is what you've read? Higher volumes would yield more accidents in total, but the accident rate for each would be the same.

Also, I think SPUIs may end up requiring even more width on the secondary road than a conventional diamond interchange (IIRC there's a DDI in Connecticut visible on the satellite view in Google Maps that seems to make this clear), whereas in some cases a diverging diamond interchange can work with a narrower road than a conventional diamond interchange, so my understanding is that DDIs tend to be a win if keeping bridge costs down is a goal.

This is absolutely correct. SPUIs require much wider overpasses or underpasses to accommodate dual left turn lanes for all ramp movements and to facilitate geometry for the central point of intersection. Here's a good example of a SPUI next to a traditional diamond: https://goo.gl/maps/zsdPBazjpLA2 - the difference between the SPUI overpass width and diamond underpass is very apparent.

But, again, in terms of DDI versus SPUI, you have to consider the traffic patterns to determine whether the increased cost of the SPUI is worth it in order to provide better throughput if all of the movements see relatively even demand.

Cloverleaf interchanges work well enough for quite a few exits. But there is a big difference between the 95/93 cloverleaf and the 95/rt 4 cloverleaf and the like. Crazy that 95/93 interchange has a pattern designed for far fewer cars.

I'd actually submit that any existing cloverleaf in MA that is not freeway-to-freeway should be converted to a partial cloverleaf or parclo.

Certain freeway-to-freeway cloverleafs need to be upgraded with flyovers (like they are doing with 93/95) or collector/distributor roadways. For everything else, a parclo would work much better than a traditional cloverleaf.

The new Route 9/128 interchange is a variant of the above parclo and should help a lot with congestion on 128 by eliminating the weaving movements inherent in the cloverleaf design.

But alas, people in MA tend to think adding lights means more traffic and are subsequently reluctant to accept any design changes to existing interchanges.

 

[HalcyonEra]

Cloverleaf interchanges work well enough for quite a few exits. But there is a big difference between the 95/93 cloverleaf and the 95/rt 4 cloverleaf and the like. Crazy that 95/93 interchange has a pattern designed for far fewer cars.

Yet we continue to fail in upgrading. How exit 17 (Rt 135) going northbound on 128/95 was not upgraded is beyond me. It is a short half cloverleaf off ramp with a very sharp turn at the bottom of a 1-2 mile hill. While I understand a diamond may not have worked due to preserved land, how there is not a deceleration lane on the approach makes no sense.

 

[tangent]

I've read SPUIs are just as safe as a traditional diamond interchange but are built to handle significantly more volume. Perhaps this is what you've read? Higher volumes would yield more accidents in total, but the accident rate for each would be the same.



This is absolutely correct. SPUIs require much wider overpasses or underpasses to accommodate dual left turn lanes for all ramp movements and to facilitate geometry for the central point of intersection. Here's a good example of a SPUI next to a traditional diamond: https://goo.gl/maps/zsdPBazjpLA2 - the difference between the SPUI overpass width and diamond underpass is very apparent.

But, again, in terms of DDI versus SPUI, you have to consider the traffic patterns to determine whether the increased cost of the SPUI is worth it in order to provide better throughput if all of the movements see relatively even demand.



I'd actually submit that any existing cloverleaf in MA that is not freeway-to-freeway should be converted to a partial cloverleaf or parclo.

Certain freeway-to-freeway cloverleafs need to be upgraded with flyovers (like they are doing with 93/95) or collector/distributor roadways. For everything else, a parclo would work much better than a traditional cloverleaf.

The new Route 9/128 interchange is a variant of the above parclo and should help a lot with congestion on 128 by eliminating the weaving movements inherent in the cloverleaf design.

But alas, people in MA tend to think adding lights means more traffic and are subsequently reluctant to accept any design changes to existing interchanges.

Would be good to see some modeling to see what volume of traffic that makes sense for. Generally it seems that if the number of cars is so low that you could replace the full cloverleaf with a partial cloverleaf then weaving cars is probably not a great concern in the first place.

Seems likely more efficient to keep the full cloverleafs where they already exist rather than incur unnecessary expense redesigning interchanges that don't have significant issues. Designed from scratch and you are probably right.

 

[mass88]

Cloverleaf interchanges work well enough for quite a few exits. But there is a big difference between the 95/93 cloverleaf and the 95/rt 4 cloverleaf and the like. Crazy that 95/93 interchange has a pattern designed for far fewer cars.

We have some of the worst highway interchanges I have seen anywhere. 93/95 in Canton and Woburn are two prime examples.

 

[tangent]

We have some of the worst highway interchanges I have seen anywhere. 93/95 in Canton and Woburn are two prime examples.

Yes, it is insane that plans to upgrade these interchanges have been hung up for so long.

 

[omaja]

Would be good to see some modeling to see what volume of traffic that makes sense for. Generally it seems that if the number of cars is so low that you could replace the full cloverleaf with a partial cloverleaf then weaving cars is probably not a great concern in the first place.

Seems likely more efficient to keep the full cloverleafs where they already exist rather than incur unnecessary expense redesigning interchanges that don't have significant issues. Designed from scratch and you are probably right.

The modeling has already been done. Safety studies have already been done. This isn't something that MA would be reinventing; we're talking about the most common interchange setup on the busiest freeway in North America: Highway 401 through Toronto area - not to mention countless other uses across the US. Parclos are far more efficient and safe for freeway-to-arterial road interchanges. So much so, in fact, that California has decided to replace existing cloverleafs with parclos on freeway-to-arterial interchanges as common practice. As interchanges come due for repaving/rebuilding, this is something MA should be considering. Improving the geometry of ramps more than makes up for the addition of lights on the arterial road.

For example, in 2007 the 128/Route 1 interchange in Dedham saw around 200,000 average daily movements. 150,000 of that was 128, so the remaining 50,000 or so is on Route 1. Look up just about any parclo interchange and you'll find that they can handle significantly more traffic than that (e.g. I-80 at 42nd, 60th, 72nd and 84th Sts in Omaha or Highway 401 at Keele St, Weston St, Dixie Road, and many others in Toronto. The limiting factor for parclos is really a combination of the number of lanes and speed of traffic on the arterial. Most interchanges in MA are already well below the vehicle count which would necessitate a completely grade-separated interchange, not to mention that travel speeds are usually already marked at 35-50 mph range, so removing two loops from a cloverleaf would not adversely affect traffic flows.

 

[Charlie_mta]

Pretty cool YouTube video on Diverging Diamond interchange in Michigan:

https://www.youtube.com/watch?v=BNGXHKpRcr8

 

[Joel N. Weber II]

Re: Cloverleaf Land Use

At any rate, yes, land tied up in the loops of cloverleafs is all wrong:
- too much extremely valuable land
- too many conflicting movements

Really bad mobility-per-acre.

If we're having a conversation about 128's interchange 19 at Highland Ave in Needham and the potential for bringing the Green Line and transit oriented development to that neighborhood, I think you have exactly the right argument.

If we're talking about the I-95/110 interchange in Amesbury, which is a partial cloverleaf but not in the good ways omaja described (110 westbound traffic receives traffic from I-95 northbound and then immediately has the ramp to I-95 southbound) and where getting rid of the I-95 northbound to 110 westbound ramp might free up space to be able to build a bike path, your argument also seems to apply.

However, most cloverleafs seem to be built in places which simply do not seem to have high density usage by humans, and perhaps it would be appropriate to think about using the ``wasted'' space for restored wetlands or solar panels.

 

[Joel N. Weber II]

Re: 128/1 interchange in Dedham

For example, in 2007 the 128/Route 1 interchange in Dedham saw around 200,000 average daily movements. 150,000 of that was 128, so the remaining 50,000 or so is on Route 1. Look up just about any parclo interchange and you'll find that they can handle significantly more traffic than that (e.g. I-80 at 42nd, 60th, 72nd and 84th Sts in Omaha or Highway 401 at Keele St, Weston St, Dixie Road, and many others in Toronto. The limiting factor for parclos is really a combination of the number of lanes and speed of traffic on the arterial.

I'd like to see a reasonably direct permanent off-road route for the East Coast Greenway from Boston to Providence eventually be developed that doesn't head to the northwest out of Boston and then through Worcester, and it seems like the route 1 bridge in Dedham might be a viable place for that route to cross 128.

A DDI would offer the option of putting the bike path in the middle of the bridge, and potentially having a bidirectional bike path on the east side of Providence Highway from 128 to the Wigwam Pond Area, and then south of 128 having the bidirectional bike path continue along the west edge of 1 for a bit before diverging from 1, possibly taking a fairly direct route from the bridge to Islington station, and then possibly continuing as rail-with-trail (or through adjacent park land, etc) along parts of the Franklin Line. With a DDI with a bike path in the middle of the road, bicyclists would only have to cross one direction of automobile traffic at a time, and I think a DDI would probably make for a less stressful configuration for bicyclists than any cloverleaf variant.

Slowing the secondary highway to 20 MPH where a DDI is expected to interact with significant bicycle activity would probably also be appropriate.

 

[Joel N. Weber II]

Re: Cloverleaf Land Use

However, most cloverleafs seem to be built in places which simply do not seem to have high density usage by humans, and perhaps it would be appropriate to think about using the ``wasted'' space for restored wetlands or solar panels.

Or maybe we could convert to wetlands, grow switchgrass, and turn that switchgrass into jet fuel (or use that fuel to generate electricity on the coldest winter days).