Stephen Rees's blog

Thoughts about the relationships between transport and the urban area it serves


with 14 comments

I have written previously in this blog about my preference for roundabouts as an intersection design. And how the intersection of Granville and Garden City in Richmond needs to be rethought.

This post cannot refer you to its source as it is only available to ITE members, but it is published in a peer reviewed journal (ITE Journal, March 2007) “A Comparative Study of the Safety Performance of Roundabouts and Traditional Intersection Controls” By Shashi S. Nambisan PhD PE and Venu Parimi EIT which was presented by the authors at the ITE 2005 Annual Meeting.

The evaluation compares the traffic crashes in the proximity of modern UK style roundabouts and intersections in Las Vegas NV, using a 5 year data set. A UK style roundabout differs from a traffic circle in that vehicles entering the roundabout must give way to traffic already in the circle. This differs from the way that US traffic circles traditionally operate, with priority given to vehicles entering the circle.

The study compares six roundabouts to eight conventional intersections. “The injury crashes at conventional intersections are significantly higher than at the roundabouts”. “Most of the crashes at the roundabouts (nearly 60 per cent) were found to be minor sideswipe collisions” “Nearly 48 percent of the the crashes at the subject STOP controlled/signalized intersections were caused by the driver’s failure to yield to traffic. Most of the crashes at the subject roundabouts were caused due to improper lane changes, inattentive driving and making improper turns.”

So, what a roundabout does is end the carnage due to red light and STOP sign running and turn these collisions into minor sideswipes. But mainly what happens is that drivers have to become active in entering the intersection, and look for a safe gap to merge into the flow around the roundabout – just as we do now when merging onto a freeway. But it happens more slowly and over a shorter distance. The “T Bone” collision we see at STOP signs and red lights is a thing of the past. Collisions occur at conventional junctions because one driver either does not notice or, more likely, deliberately flouts the rules. One driver sees a yellow light and speeds up: the other driver sees a green light so decides to proceed without checking to see if the intersection is clear. Bang.

At a roundabout the sign is clear: it says “Yield”. You have to stop and look before proceeding. If there is a collision it’s because someone isn’t paying attention or is pushing his luck. Either way, the outcome is less likely to be injury or a very expensive. Roundabouts do reduce collisions, but they still occur, they just don’t cause as much damage. A good principle to apply to most urban traffic management .

So here, courtesy of Google Maps and Windows Paint, is my free for the taking redesign of the worst intersection in Richmond.


And for those who need references

Federal Highway Administration – Safety Brief

Information guide

TRB Modern Roundabout Design

and the second installment added May 31

Written by Stephen Rees

April 17, 2007 at 5:16 pm

14 Responses

Subscribe to comments with RSS.

  1. For that particular intersection, that roundabout design look even more confusing than what’s there. i.e. does southbound->westbound traffic even enter the roundabout at all? Likewise, would eastbound traffic on Granville enter the roundabout?

    Maybe if the intersection were normalized into more of a cross configuration with the roundabout, that would be more predictable for drivers to navigate. (i.e. sharpen the big swooping curve)

    BTW, what do you think of the City of Vancouver’s new traffic signals at Alma and Point Grey Road, which deals with a similar configuration?


    April 19, 2007 at 2:17 pm

  2. All traffic enters the roundabout. It is a very simple four leg configuration. I did not try to move anything needlessly – and using Microsoft Paint one is compelled to be very simple. I suspect that a proper geometric design program could save a lot of space at this junction.

    Each entry arm has two lanes. Left for traffic going around the centre, right for leaving at the first exit. The circulating area is similarly two lanes wide, but has no lane markings. The only paint on the road are the lanes on the entrances exits and the yield line, supplemented by inverted red triangle sign posts at each yield point. The number of potential conflict areas is reduced to four.

    I have not see the intersection you refer to but all the other circular junctions I have seen in Vancouver are US style traffic circles which are useless and widely flouted by people driving wrong way round to make a left.

    Stephen Rees

    April 19, 2007 at 3:21 pm

  3. The Richmond Review piece that quotes me also has an explanation about the layout.

    Its original design can be blamed on the interurban tram that once connected Steveston with downtown Vancouver.

    The rail corridor included Garden City Road and Granville Avenue. Since trains can’t make sharp right turns, the track slowly curved—and the road, built in the ‘60s or ‘70s, followed suit.

    for a map see

    Stephen Rees

    May 3, 2007 at 9:21 am

  4. I saw that article today – interesting tidbit.
    The Alma / Point Grey Road in Vancouver used to be a through movement for one of the turns (the double yellow centre line followed the “L” of the turn, and there were stop signs for the other two arms entering the intersection. Now the interesection is fully signalized, with priority in signal timing given to the major “L” movement and probably sensors for the other two arms.
    Before the signal was installed, drivers would fail to signal a turn when taking the turn, because they were following the yellow line – that made it difficult for pedestrians and drivers at the stop signs to predict what the other driver was doing.


    May 3, 2007 at 10:08 pm

  5. I spoke to Victor Wei (the city’s director of transportation) today. He told me that he has already has received a number of calls from people who oppose the idea of a roundabout. The reason they give is that Richmond drivers are thought to be some of the worst in the region, and will not know what to do when faced with a yield sign at a roundabout. Which is interesting since the City has already put yield signs on a traffic circle on Saunders Road (between No 3 and Garden City) in an attempt to impose better discipline on local traffic. Saunders is one of those rare straight through residential streets that are not part of the arterial grid, and thus popular as “short cut”.

    The “urban legend” of bad drivers in Richmond has some basis in fact. An unknown number of Richmond residents bribed their way through the driving test. Though the examiner and a driving school owner were caught and penalized, none of the drivers were required to retake their test.

    Stephen Rees

    May 9, 2007 at 6:34 pm

  6. Dear Mr. Rees:
    Thank you for your comments regarding traffic circles in Richmond. Your comments have been forwarded to Victor Wei, Director of Transportation for review and response.
    Transportation Division
    City of Richmond

  7. Thank you. They are not “traffic circles”, they are roundabouts. Quite different.

    Stephen Rees

    May 31, 2007 at 3:33 pm

  8. Hey Stephen, are you still on about this or did your page vent your rant? I am in San Antonio and been going on about traffic flow for years (I used to live in P’boro England) I get the same responses: “Nobody will know what to do.” “It’s confusing.” And worse, “Roundabouts cause more accidents.” The part that really gets me going is that most of these responses come from CITY ENGINEERS, the very people that don’t read the actual statistics they claim to be using to design roads!

    Have you made any headway in your area?

    PS Loved your response to the Richmond Transportation Division wonk.



    August 8, 2007 at 9:54 am

  9. There are, in fact many roundabouts in Grater Vancouver – my most recent find was in Yaletown. ICBC – our provincial agency for mandatory vehicle insurance has been promoting them here for years. It is uphill work. No changes have been made yet to the one illustrated in the op.

    Stephen Rees

    August 8, 2007 at 12:16 pm

  10. […] Roundabouts have been covered extensively in this blog, because while Britain has been using them for years to reduce collisions at intersections, the “not invented here” syndrome has kept them out of most North American cities. ICBC actively promoted them for some years – and has an impressive data set to back up the experience. But talk to many City engineers and they will be dismissive – and, mostly, appallingly ignorant. And will repeat twaddle like “drivers here will never understand them”, which always prompts me to ask what is so different to drivers “here” than North Vancouver, or Agassiz, or Pemberton? At intersections with stop signs or traffic lights, the most common — and serious — accidents are right-angle, left-turn, or head-on collisions that can be severe because vehicles may be moving fast. Roundabout virtually eliminate those types of crashes because vehicles all travel in the same direction. […]

  11. Stephen,

    I finally got around to writing my own blog on the subject. You might find some of the links useful.



    September 27, 2008 at 2:04 pm

  12. as a Canadian having lived around the world, including in Australia where roundabouts are de rigeur, i find them an excellent, efficient means of traffic control. that said, people elsewhere know how to use them, because they’ve been trained by observation growing up and of course by training. apparently not so in Vancouver.

    i can’t count how many times i’ve had to brake hard, while IN a traffic circle (ie, just in front of the person to my right trying to enter), because they are thinking that i’m supposed to yield to them. meanwhile, in the cascade of rules that i understand, and that seem completely logical in terms of flow, the person already in the circle – whether big roundabout or little traffic calmer – should have the right of way. why the difference in rules between “roundabouts” (such as can be found just south of UBC, or on the last exit into White Rock, with those great big Yield signs) and “traffic circles” (such as the little traffic calming circles around Kits)? no wonder people here don’t know how to use them.


    October 7, 2008 at 1:21 am

  13. Some of you may not be aware that I commented (quite negatively) on Nambisan and Parimi’s ITE Journal article, based on a two day site visit to all the locations in Las Vegas, mentioned in the article. My “Letter to the Editor” was published in its entirety in the June 2007 edition of ITE Journal. Normally, authors provide a rebuttal, which would then also be published. To date (July 2010) this rebuttal has not been seen. This shows me that the authors likely cannot disagree with my observations. The saddest part is that this paper was already presented in Australia. A lack of peer review?

    Here is my letter in the June 2007 ITE Journal:


    July 8, 2010 at 9:29 pm

  14. A serious accident at a roundabout-under-construction in the Township of Langley (232rd Street and 56th Avenue) hit the headlines today. Perhaps some investigation might be worth it.


    May 29, 2012 at 9:56 pm

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: