4-way STOP v. Roundabout efficiency - by the Mythbusters

[mgk920]

Even these guys' simple test is an eye-opener. And they did not even measure the time delays of vehicles waiting in line to go through the 4-way STOP v. the roundabout.

Enjoy!

Mike

[hotdogPi]

A roundabout is usually instead of a traffic light, not instead of a 4-way stop.

(I will not spoil the final result of the video.)

[corco]

A roundabout is usually instead of a traffic light, not instead of a 4-way stop.

That's...a bit of a generalization

[NE2]

A roundabout is a 4-way yield, so of course it's more efficient than a 4-way stop.

[empirestate]

I think the real myth is that Americans have little exposure to or experience with roundabouts. Seems like every week for the past 10 or 15 years now, a new roundabout has opened somewhere in the U.S. with an accompanying news article about how confused drivers are by the technique. Confused they may indeed be, but I think the reason for that can no longer be that roundabouts are new or rare in this country.

[mass_citizen]

A roundabout is usually instead of a traffic light, not instead of a 4-way stop.

That's...a bit of a generalization

Massachusetts is home to a number of roundabouts and even more of its bigger cousin, the dreaded rotary. In my experience usually a light takes the place of a rotary or a roundabout takes the place of a light. I do not know of any instances of using a roundabout instead of a 4 way stop. In such a case of upgrading a 4-way however, I'm sure the delays and level of service provided by a light vs. a roundabout would be analyzed. Usually, the roundabout wins.

[corco]

Right, in Arizona though where I lived recently, roundabouts were always used in place of four way stops where traffic volume wouldn't warrant a stoplight, like so http://goo.gl/maps/WExDo.  I know of one exception:

http://goo.gl/maps/Uj9lV

Actually, I could be wrong on this- when I think of all the roundabouts on SR 89/89A/179 up around Sedona and Prescott, those are often used in lieu of stoplights.

Other states I've lived in...I think Washington had some in lieu of stoplights but also some in lieu of four way stops, there aren't enough of them that I've seen in Montana or Wyoming to really tell, Idaho had a few in lieu of four way stops.

[hbelkins]

There's going to be a roundabout at the intersection of US 62 and the new US 27 bypass of Cynthiana, Ky. The intersection is in somewhat of a rural area. I suspect the roundabout may have been installed in lieu of stop signs on US 62.

[andy]

Roundabouts are ok in some places, but it seems they need more real estate and don't always mix well with pedestrians.

In Indiana, it seems north of Indianapolis and Bloomington are fairly aggressive with using them.

[Big John]

Because of traffic counts dictating widths, 1-lane roundabouts are often used in lieu of a 4-way stop while roundabouts with 2 or more lanes are usually used in lieu of a traffic signal.

[Joe The Dragon]

What about Roundabouts vs clover leafs? in some places the UK over does Roundabouts where they need more grade separation

[realjd]

Roundabouts are ok in some places, but it seems they need more real estate and don't always mix well with pedestrians.

In Indiana, it seems north of Indianapolis and Bloomington are fairly aggressive with using them.

Roundabouts don't have to be small. See the miniroundabouts that are common in the UK. It's basically a painted circle in the middle of a small intersection and forces drivers to use roundabout rules (yield to cars on the right) at what would be a 4-way stop in the US.

[andy]

Roundabouts are ok in some places, but it seems they need more real estate and don't always mix well with pedestrians.

In Indiana, it seems north of Indianapolis and Bloomington are fairly aggressive with using them.

Roundabouts don't have to be small. See the miniroundabouts that are common in the UK. It's basically a painted circle in the middle of a small intersection and forces drivers to use roundabout rules (yield to cars on the right) at what would be a 4-way stop in the US.

I think you meant "can be small".  Would that be much different than some type of 4 way yield, to which I would not be opposed.  My apologies to law enforcement, but how much gas and time could be saved if rolling stops (where reasonable and safe) were permitted.

[Indyroads]

I am still not convinced that roundabouts are a good substitute for a very busy intersection with high traffic volume. I think that traffic signals with a proper traffic flow control system (ITS) are more efficient in areas where volume is high. In Carmel Indiana it works ok at "MOST" of the roundabouts they have put in. but there are some glaring failures where they roundabout has caused traffic backups. Such as the closely spaced roundabouts at 116th and Springmill and 116th and Illinois which are located west of US-31 https://maps.google.com/maps?q=116th+and+meridian&rlz=1C1AVSX_enUS390US395&um=1&ie=UTF-8&hl=en&sa=N&tab=wl If we were to put these kind of roundabouts in at Allisonville and 82nd street or Michigan and 86th street they would be likely disastrous to traffic flow during peak periods. Same as in Citrus Heights CA. if we replaced the signals along Sunrise Blvd from Madison all the way to Roseville CA, there would be serious traffic jams at Madison Ave, Greenback Lane, Old Auburn, Antelope Road, etc. as well as bumper to bumper traffic along Sunrise Blvd. a street that is usually very heavily traveled most all hours throughout the daytime.

[J N Winkler]

A roundabout is usually instead of a traffic light, not instead of a 4-way stop.

That's...a bit of a generalization

It is, but there is a very wide AADT range for which roundabouts and traffic signals are essentially interchangeable as intersection treatments, with stoplights having an advantage for unbalanced flows while roundabouts generally lead to less frequent and less severe side-impact crashes.  Four-way stops, on the other hand, are able to cater only for a narrow AADT range at the low end.

Perhaps a more precise way of expressing the distinction is to say that while roundabouts can replace traffic lights and four-way stops in some circumstances, a four-way stop is unlikely to work well as a substitute either for traffic lights or a roundabout.

Other states I've lived in...I think Washington had some in lieu of stoplights but also some in lieu of four way stops, there aren't enough of them that I've seen in Montana or Wyoming to really tell, Idaho had a few in lieu of four way stops.

It's been ten years since I was last in Washington state (other than to change planes), but I don't remember actually seeing any roundabouts deployed at intersections where the flows were blatantly too heavy for a four-way stop.  This was the only roundabout I personally drove when I was last in Washington state, and traffic flows seemed light enough (at least in off-peak) to be handled by a four-way stop:

I-5 off-ramp at SR 510

This, on the other hand, seems to handle flows heavy enough to justify a traffic signal as an alternate to the roundabout:

SR 510 at Pacific Ave. SE and Marvin Way

In Kansas I have never personally seen a roundabout used at a location where the traffic volumes were heavy enough to justify a signal.  KDOT has typically used roundabouts in the following scenarios:

*  Rural highway intersections with bad accident records--typically as a replacement for stop control (which handles the volumes adequately if not in complete safety)

*  As part of diamond interchanges in rural areas (e.g., US 75 north of Topeka) or in small towns (e.g., I-135 in Newton), generally at locations previously handled with stop control

Roundabouts are also used in urban areas in Kansas.  The roundabouts I am familiar with in Wichita, Topeka, and Manhattan are all in neighborhoods or on rural arterials where stop control (either two- or four-way) would have been used as an alternative.

There was a proposal to include a roundabout in the Turnpike/Kellogg/Webb interchange reconstruction in Wichita, but it was shot down in part because the design was considered too experimental for the heavy flows that needed to be accommodated.  So although Kansas was one of the early pioneers in roundabout development, I sense a reluctance to use them for heavy flows that is only just now starting to fade.  Kansas is nowhere near as aggressive in this regard as, say, Wisconsin, where US 41 is being built with multiple interchanges where multi-lane roundabouts are designed to handle quite heavy traffic volumes and have overhead signing to provide added positive guidance (similar in function if not appearance to the Queensland MUTCD-style multi-lane roundabout diagrammatics used around Prescott in Arizona).

I think the real myth is that Americans have little exposure to or experience with roundabouts. Seems like every week for the past 10 or 15 years now, a new roundabout has opened somewhere in the U.S. with an accompanying news article about how confused drivers are by the technique. Confused they may indeed be, but I think the reason for that can no longer be that roundabouts are new or rare in this country.

Considered at the national level, roundabouts are not rare in any absolute sense, nor are they novel any longer.  But there is still very broad regional variation both in how frequently they are used overall, and how frequently they are used for heavy flows.  Wisconsin probably represents one extreme:  roundabouts used for preference (signals can be used only if specially justified), and roundabouts routinely used at very busy intersections.  At the other extreme you have agencies like TxDOT:  I can count on the fingers of one hand the number of contracts for roundabouts on the Texas state highway system I have seen in the last ten years (extent of the system is close to 80,000 miles and there are probably about 1000 contract advertisements annually).  Most other state DOTs, including Kansas DOT--which was one of the early adopters and is very active in research on how to adapt roundabouts for overweight/overdimensional loads owing to the heavy wind industry in Kansas--are somewhere in the middle.

So, put simply, I don't think roundabouts are anywhere near "finding their level" as an intersection treatment in the US as a whole, so it doesn't really surprise me that they still cause localized confusion more than 20 years after the early experiments in Summerlin.

[roadfro]

Having taken traffic engineering classes and calculated delay and capacity for all-way stops and roundabouts (by hand, no less), I can confidently say the Mythbusters' findings are valid. I think this explains the benefit of roundabouts over stop control better than formulas and data ever could.

What would have been even more interesting to see is whether their results would have been much different if they used a 2-lane setup...



Nevada has embraced the roundabout concept and has tried them in a number of urban and rural locations. There have been very few intersections where one hasn't improved traffic flow, and this goes for installations that have replaced both all-way stops as well as traffic signals. It has become pretty standard practice by NDOT and several local agencies to evaluate a roundabout as a potential option when looking at intersection improvements--however, they aren't promoting or installing them nearly as aggressively as, say Wisconsin has along the US 41 corridor.

So, put simply, I don't think roundabouts are anywhere near "finding their level" as an intersection treatment in the US as a whole, so it doesn't really surprise me that they still cause localized confusion more than 20 years after the early experiments in Summerlin.

For those that don't know, 'Summerlin' is a master-planned community in Las Vegas, NV, which is home to the first two "modern" roundabouts in the United States. I believe they were installed in 1990. Since the initial installation, Las Vegas (and the Summerlin area in particular) have had a number of roundabouts installed.

Town Center Drive & Village Center Circle
This is the biggest of those initial two roundabouts. Town Center is one of the major north/south streets in Summerlin, departing to the southwest--it has a freeway interchange about 1/4 from this location. The physical characteristics of this roundabout have not changed since original construction (to my knowledge), but striping has changed for additional guidance. I think it started off as partially 3-lane circulation (it may not have had any internal markings when it first opened) but now operates as a two-lane.

Town Center Drive & Hualapai Way
This is a later roundabout in Summerlin about 1 mile southwest of the previous one. I think it was built in the early 2000s. Both Town Center and Hualapai are 3-lane arterials on all legs of this, but the right lane on each approach is a dedicated right turn separated from the circular roadway so it actually functions as a 2-lane roundabout. This intersection was built as a roundabout at initial construction, but would likely have been a signalized intersection if built in any other part of town. It operates very effectively (note that there is a hospital in the northern quadrant, so a significant traffic generator is nearby). Part of this is attributable to much more adequate signage and lane markings providing positive guidance to drivers.

[doorknob60]

As someone from Bend, OR (practically the king of roundabouts in the US, or at least the western states), roundabouts are definitely more efficient than 4 way stops in most situations. I like them.

[mgk920]

What struck me was the observation that was noted in the clip that many more vehicles were in the intersection at the same time, as many as six, with the roundabout v. the 4-way STOP, meaning that they are likely getting through the intersection much more quickly.

Mike

[froggie]

Couldn't see the vid, of course (#LifeAtSea), but judging from the comments, sounds like it reinforces the concept that roundabouts are efficient.

[vdeane]

I'd be interested to know what the results of a traffic light would be with their tests, especially since an intersection handling that many cars would have one in my area.

[Indyroads]

I'd be interested to know what the results of a traffic light would be with their tests, especially since an intersection handling that many cars would have one in my area.

that depends on the traffic signal. Some signals are set so stupid that you have to wait 1-2 minutes while no cross traffic is passing by for a green light. this is typical with timed signals as opposed to video/loop detection.

[vdeane]

I'd be interested to know what the results of a traffic light would be with their tests, especially since an intersection handling that many cars would have one in my area.

that depends on the traffic signal. Some signals are set so stupid that you have to wait 1-2 minutes while no cross traffic is passing by for a green light. this is typical with timed signals as opposed to video/loop detection.

That is also typical of every single signal in my city.

[tradephoric]

I'd be interested to know what the results of a traffic light would be with their tests, especially since an intersection handling that many cars would have one in my area.

that depends on the traffic signal. Some signals are set so stupid that you have to wait 1-2 minutes while no cross traffic is passing by for a green light. this is typical with timed signals as opposed to video/loop detection.

An adaptive traffic signal can still be quite inefficient.  Consider a 4-phased intersection.  Each phase is going to have roughly 5 seconds of wasted time  (~3 seconds wasted time for safety clearances and ~2 seconds wasted start up time).  A 4-phased intersection that is running a 100 second dial has roughly 20% of wasted time where nobody is moving through the intersection. 

In addition, a side-street pedestrian phase can greatly reduce the efficiency of an adaptive intersection.  A side-street pedestrian crossing of 150 feet (which is not uncommon) would require 50 seconds to allow the pedestrian to safely cross.  Due to the pedestrian clearance intervals, a side-street that may only require 15-20 seconds to clear out the vehicle traffic would still need to run a full 50 seconds to fit the peds.  Adaptive signals aren't really that adaptive when you consider all the safety minimums that must be satisfied each cycle.  An adaptive system also requires accurate working detection which isn't always the case (loops break, inaccurate traffic counts with video detection, broken push buttons, etc. etc.).

[jeffandnicole]

I went thru this intersection/roundabout recently. And as luck would have it, the aerial photo shows the older configuration, and GSV shows the newer configuration.

This is Spring Valley Road, which goes from US 322 to US 202 in the Concordville, PA area.

The old intersection, which was a 4 way stop: http://goo.gl/maps/VlFS8

The new roundabout...with a 4 way stop! http://goo.gl/maps/EQv15

[1995hoo]

I love roundabouts. My biggest beef with American roundabouts is with drivers who fail to signal. If signals are used properly, other drivers need never guess at what you will do, such that traffic moves more efficiently. If you fail to signal, it freezes entering drivers who don't know whether you plan to exit and so have to freeze in case you stay on the roundabout.