Dec 05

More troubling statistics on Oregon’s dual-lane roundabouts

Washington County has finally announced a public open house to discuss their plans to “improve” Highway 47 on the northern periphery of Forest Grove, and the related David Hill Road extension.  It’ll take place on Thursday, December 11th from 5-7pm, in the Forest Grove Community Auditorium (1915 Main Street).  The official announcement itself is here.

There will be no formal presentation, but county staff (to whom ODOT has outsourced a substantial amount of project management) will be there to show plans and take feedback.  They also finally admitted publicly that they do indeed plan to build not one but two roundabouts on this section–the previously announced one at the Verboort/Purdin intersection, and, confirming the information I broke here, a second one where Highway 47 will meet the new David Hill Road extension.

In that same analysis, I broke down the statistics for the last dual-lane roundabout that ODOT built on a state highway: the one at the base of the Youngs Bay Bridge in Astoria, where US Highway 101 meets Oregon Highway 202.  While ODOT spokesman Lou Torres claimed that this roundabout “functions well“, the actual long-term crash data showed it does not.  Comparing “before” and “after” periods of equal length, I showed that this roundabout actually caused a 125% increase in injury accidents, without a reduction in the severity of those injuries.  This data stands in sharp contrast to Mr. Torres’ comments, as well as to the various roundabout studies out there, which, as I noted before, have been concerned mostly with single-lane roundabouts, rather than dual-lane ones like the one in Astoria, and the ones proposed for Highway 47.

However, being the policy nerd that I am, I couldn’t leave it at that, and I continued to pour over ODOT’s statistics for roundabouts and signalized intersections.  At the April town hall meeting that was prompted by the last fatality that occurred at the Verboort/Purdin intersection, ODOT Region 2 Manager Sonny Chickering spoke highly of the dual-lane roundabout that his hometown of Springfield built in 2006.  That roundabout exists where Pioneer Parkway, Martin Luther King Jr. Parkway, Harlow Road, and Hayden Bridge Way intersect, near the Gateway Mall and Sacred Heart RiverBend Medical Center.  The speed limits on the approaches to this intersection are higher than the Astoria roundabout: Harlow and Hayden Bridge have 35mph speed limits, while Pioneer and MLK are posted at 45mph.  Per p. 2 of the brochure that the City of Springfield produced about the project, the matter of funding is discussed.  Per the listed figures, the roundabout received $9.34 million worth of funding, obtained from Lane County and PeaceHealth, plus unspecified “contributions from the City [sic]“.

City of Springfield Diagram of Dual-Lane Roundabout

City of Springfield Diagram of Dual-Lane Roundabout

How well does this $9.34 million+ roundabout function from a safety standpoint?  Because of the fact that this intersection has major roads with four different names entering it–and Pioneer Pkwy is farther divided into “West” and “East”–one has to do a search for each leg of the roundabout in order to get the full statistics from ODOT’s Crash Data System.  For the purposes of this analysis, I will use the 5-year period from 01/01/2009 (the earliest date for which data is available) to 12/31/2013.

Some of the points on the roundabout, such as Pioneer Pkwy W and Harlow Road, give off no reports, which, initially, will make this roundabout seem miraculously safe.  However, if one tries the other legs, a completely different picture emerges–see Pioneer Pkwy W and Hayden Bridge Way, Pioneer Pkwy E and Hayden Bridge Way, and Martin L King Pkwy and Hayden Bridge Way.  If we tally the results of these three documents, we now have a more accurate picture.  For a refresher, here’s how ODOT classifies accident severity:


Code Short Description Long Description
K Fatal injury
A Incapacitating Injury/Major Injury “Prevents person from walking, includes severe lacerations, broken limbs, abdominal injuries.”
B Non-Incapacitating Evident Injury/Moderate Injury “Evident to observers, lump on head, bruises, cuts.”
C Possible Injury/Minor Injury “Limping, momentary unconsciousness.”
O No Injury/Property Damage Only (sometimes also referred to by the acronym “PDO”)

And if we tally up the results from these three legs, we get the following results:

Approach K A B C O Total Crashes Total Injury Crashes Total Injuries
Pioneer Pkwy W/
Hayden Bridge Way
0 0 2 10 34 46 12 17
Pioneer Pkwy E/
Hayden Bridge Way
0 0 3 17 58 78 20 27
Martin L King Pkwy/
Hayden Bridge Way
0 0 2 13 42 57 15 20
TOTAL 0 0 7 40 134 181 47 64

Note that underlined cell on the above table. Per the available data, it appears that the Springfield roundabout has been the site of 181 accidents in just 5 years, which injured 64 people. The one positive is that there were no fatalities or Class A injuries, but the rest of these results suggest that this roundabout experiences an absurdly high accident rate. 181 accidents in 5 years, regardless of severity, is an astonishingly terrible number. These results, coupled with the Astoria results, should give engineers serious pause before even thinking about building more dual-lane roundabouts in Oregon.

The City of Springfield does not appear to have up-to-date traffic count information available, so it is difficult to judge the actual volume this roundabout presently receives–based on this 2008 map, the Pioneer Pkwy approach is busiest. Adding the two sides of Pioneer Pkwy together, it would appear the AADT just south of the roundabout was 19,413 at that time. Fortunately, the City of Springfield’s 2035 Transportation System Plan (TSP) does give the crash rate in the Million Entering Vehicles (MEV) figure that many traffic engineers use, and the data for the Springfield roundabout (identified as Pioneer Parkway/Hayden Bridge Way). The information can be found on p. 31 of the PDF for Volume 3, Appendix B. Here’s a screencap of the relevant table (click to enlarge):


Table 7, Vol. 3 App. B of Springfield 2035 Transportation System Plan.

As you can see by the above table, the Springfield dual-lane roundabout has a crash rate of 2.91 MEV, nearly three times the crash rate of the next worst intersection studied. (A footnote on the next page clarifies that the data for the roundabout is actually collected from the period of October 2006, when the roundabout was completed, through the end of 2009, rather than the full 2005-2009 figure used for the other intersections.)

According to Chapter 4 of the ODOT Analysis Procedures Model document (p. 2), in all its awkwardly-worded glory, “[t]he old rule of thumb wasthat intersections with a crash rate of 1.0 per MEV or greater is generally considered to be an indication that further investigation is warranted.”  The document manual then goes on to conclude that this “old rule of thumb” should no longer be used, because it “led to ignoring of safety issues when the crash rate was below 1.0”.  In other words, a figure of 2.91 MEV, under current standards, is even worse than when the “old rule of thumb” was applied.

For an amusing comparison, here’s the 2009-2013 statistics for a major signalized intersection in the state.

K A B C O Total Crashes Total Injury Crashes Total Injuries
0 0 3 18 25 46 21 26

What is this mystery intersection?  Hint: It’s in Hillsboro.


185th and Evergreen: statistically safer than Springfield’s dual-lane roundabout, and it’s not even close.

It’s 185th Ave and Evergreen Pkwy.  Despite the fact that the segment of 185th north of the intersection has a traffic volume of more than 54,000 AADT–2.5 times the most traveled approach of the Springfield roundabout–there were only 46 accidents there during the same time period, the most severe of which were Class B, and there were less than half as many of them.  The actual data sheet is here, if you want to suspend your disbelief.

Quite simply, Springfield’s Pioneer/Hayden Bridge roundabout, in every sense of the term, a demolition derby.  And it’s demolition derby that came with a nearly $10 million price tag, to boot.  Anyone who would dare recommend multi-lane roundabouts based off the data from Springfield and Astoria is, to be blunt, simply delusional.  I still believe that the proper solution for this corridor is, at least in the interim, signalize the Verboort/Purdin and David Hill intersections with Highway 47, to address the immediate, pressing safety concerns.  This will prevent another heartbreaking tragedy, while all the stakeholders–ODOT, the county, and the cities of Forest Grove and Banks–to identify a long-term solution that will address safety and capacity issues, and dovetail with our area’s needs for more east-west connectivity.  For the amount of money that the roundabouts are likely to cost, ODOT could save up and build a proper interchange for this vital corridor.

I urge everyone in Forest Grove and the surrounding area to attend the open house this coming Thursday night.  Tell Washington County’s representatives about the alarming stats in Astoria and Springfield, and that you want to see a real solution, not an expensive and risky halfhearted attempt at one.