Traffic estimator flawed, formula being adjusted

Posted: November 30, 2013

The Winchester Star

WINCHESTER — The recent discovery of flaws in a local traffic model has not affected any roadway projects, a Virginia Department of Transportation official said this week.

Terry Short, planning manager for VDOT’s Staunton District, said the traffic-demand model in question was created by department staff members for the Winchester-Frederick County Metropolitan Planning Organization (MPO) upon its formation in 2003.

But population and growth estimates used in that model have not been realized, he said. Over the last decade, the gap between the “overly optimistic” projections and reality has grown so large that the data from the model are skewed.

“Traffic-demand models are very human: They’re only as good as the inputs that go into them,” Short said. “The model is fine. It’s the inputs that went into it that we have to fix.”

To correct the problem so the model can be used to help plan four upcoming projects, the MPO’s governing Policy Board has voted to hire a contractor to update the data.

The MPO region is comprised of Winchester and urbanized areas of Frederick County, including Stephens City. The organization’s mission is to encourage regional planning solutions for the area’s transportation needs.

Short said a consultant was beginning work on a project to design a new Interstate 81 Exit 313 interchange when the error was discovered. Projected future volumes were far greater than expected.

“The volumes were pretty extreme that it was kicking out, so it was quite clear that they weren’t particularly realistic for the period we’re looking at now,” said John Bishop, Frederick County’s deputy planning director for transportation, who was consulted by VDOT about the problem and backed the solution.

Tim Youmans, Winchester’s planning director, agreed that it was obvious that the volume projections were incorrect.

“Especially for [Front Royal Pike/U.S. 522], the projections were way beyond what the physical capacity of the roadway was,” he said. “It wasn’t so much alarming as it was a realization that there’s something that’s not working correctly in the model.”

If they had not been caught, Short said, the higher projections could have resulted in roadways designed to handle more traffic than necessary. That could have led VDOT to spend more money for right-of-way acquisition and construction; nothing had been spent on either.

“Unfortunately, we had a hiccup,” he said, “but it should be a very brief hiccup. We’ve told our contractors that we should have something in six months at the latest.”

Short said the model uses projections for population and the number and location of households and workplaces to determine how people are expected to travel on local roadways in the future.

The projection assumptions, he said, were commonly used around the country when the model was created. Adjustments were made in 2007 and 2011, but a comprehensive evaluation of the inputs has not been conducted since 2003.

Short said the problem was not discovered sooner because the model has not been used to help design any projects since the model was created.

He and Youmans noted that the MPO was planning to update the model within the next couple of years as part of the process of revising the area’s Long-Range Transportation Plan, so the project just moves up the timeframe for work already on the horizon.

Once the model is repaired, it should get quite a workout.

In addition to the Exit 313 project, Short said, the model is expected to be used on efforts to revamp the I-81 Exit 317 (Martinsburg Pike/U.S. 11 north of Winchester) interchange, help to plan the Va. 37 eastern bypass, and seek ways to improve traffic flow in and around the city’s southern tip (the Valley Avenue-Monticello Street-Shawnee Drive area).

— Contact Vic Bradshaw