Friday, September 30, 2005

Hey, It's Better than Alaska's Bridge to Nowhere!

I'm not a fan of Congressional pork barrel, but the $40 million "traffic calming" initiative for Rt. 50 in Loudoun County -- 80 percent of which will be funded by a federal earmark -- may provide a useful demonstration effect on the value of traffic circles. The initiative, a pet project of U.S. Sen. John Warner, would install traffic circles at three congested intersections along the scenic byway.

Computer models show that a traffic circle would improve the level of service at Gilbert’s Corner from an F, which is failing, to a B, the second-highest rating. Movement through the roundabout will be 10 to 15 mph slower than traffic along the main roadways, but cars keep moving, unlike at a stoplight, where drivers have to come to a stop and wait for the light to change.

“Roundabouts are used all over Europe and are growing in use in the U.S.,” said Supervisor Jim Burton (I-Blue Ridge). Besides reducing congestion, traffic circles nearly eliminate fatalities because there is no opportunity to “t-bone” someone at full speed. “The series of roundabouts for the area will greatly improve the safety.”

Backers of the roundabouts note that they will alleviate congestion on Rt. 50 for a fraction of the cost of widening the road. Read more at Leesburg2Day.


At 8:17 AM, Blogger Ray Hyde said...

In New England traffic circles are still in use. In some cases they work well other times they don't. A critical feature is understanding who has the right of way and taking turns on the merges: in this area that may take some education.

Traffic circles seem to work well when the circle itself is aq single lane. Where circles are more than one lane deep, like the infamous Bourne roundabout they are a fender bending wonder. What happens is that people turning left go threequarters around the circle and gravitate to the inside lane. Then when their exit comes they merge right and across the outside lane.

This might be a problem at Gilberts Corner because a lot of people going westbound on 50 want to go souuth on 15.

At 9:37 AM, Blogger JHop said...

Using a multi-lane roundabout does take some getting used to. The lane drift that Ray Hyde describes may be the driver's natural inclination. But it didn't take me long, using the roundabout at Old Cutler Road in Miami, to learn that I needed to place myself within the circulating traffic according to where I needed to emerge. I think they're an improvement on the 90-degree intersection, and still more useful when geography prevents the roads from meeting at right angles.

At 4:16 PM, Blogger Ray Hyde said...

Fortunately the plan for Gilbert's corner does indeed include a one lane roundabout.

At 4:11 PM, Blogger Ray Hyde said...

It is interesting what you say about level of serice. Interstates are a special case because they need high speeds and lots of open space between vehicles to be effective and safe. (Never mind that drivers tend to clump up into platoons that travel at ridiculously unsafe intervals).

At the other end of the spectrum you have Gilberts Corner or the intersection of Sudley Road and Rte 29 by the Battlefield or the intersection of Rte 28 and Linton Hall rd which are all service level F. What service level F means that not only is it bad, but it is unpredictable, because when service breaks down the queue goes to infinity.

But except for the extremes of failing roads or intersections (which are really the limiting factor, and also the major expense) what you would like to see for roads is service level D.

That is the level at which a road provides the highest throughput of vehicles, and hence the most vehicle miles for your investment in roads. Unfortunately the speed at this level of service is only 25 to 35 mph. At this speed the road is providing maximum return on investment, in terms of throughput.

The cars are not, however, because they are designed for a level of service higher thatn even interstates can provide. Cars are not designed for 35 mph, and it is too bad.

The problem with service D is that it quickly turns to level F if there is any kind of problem. So maybe you would really like to have Service level C to get good use of your roads and have some cushion against bad events. You still only need to design cars for 45 mph to get this level of service.

It is obviously unreasonable to expect that we could ever achieve uniform distribution and design of vehicles in order to get maximumum use out of our road investment, and energy and safety resources, but it is a nice idea. You have to have some slack in the system, which can also be viewed as waste or low usage level or poor investment, in order for the system to work well. The tighter you pack it, the more interactions you have, the more friction, and even more waste in terms of energy - and frustration.

But looking for this uniform level of service for all our roads would require, guess what? It would require that we have perfect control over our land use, choice of homes and job distribution, and freight and goods delivery of all sorts. If you think that uniform road usage is unreasonable or unrealistic, then you can see that optimum land use is also unrealistic: you have to have some slack or unused space in the system for it to work efficiently, otherwise you wind up with $2 million condos, like in Aspen.


Post a Comment

<< Home