By Bryant Ficek, PE, PTOE
Most, and should be all, traffic engineers will readily admit our standard process for preparing Traffic Impact Studies has several points of assumptions, estimates, or flat-out guesses.
Here are the Top 6 Areas a reviewer could focus on to pick apart a Traffic Impact Study:
- Traffic Counts – we generally complete intersection turning movement counts on one day and then use that snapshot as the foundation for all of our analyses
- Trip Generation – we use the national ITE average rate for each land use, which ignores potential high/low outliers and doesn’t account for local variation or rapidly changing traffic trends
- Modal Split – many times ignored or a general reduction applied to account for transit and other modes of transportation (ITE has recently attempted to correct this)
- Pass-By/Multi-Use/Internal Trips – ITE again provides a national basis for some land uses, but the dataset is very limited
- Trip Distribution – generally based on existing travel patterns (from the intersection turning movement counts) which would not account for regional growth, development patterns in an adjacent city, or other similar types of factors
- General Background Growth – to account for non-specific growth in traffic, a percent increase is usually applied to the existing volumes, sometimes based on historic growth or a regional model
If taken to court, would we have a better defense for the above assumptions than we used “engineering judgment?”
Our tools for traffic studies have significantly improved through camera data collection and analysis programs – among others. Yet our basic process for preparing Traffic Impact Studies (and other traffic studies) has been the same for my twenty year career.
Mike and I are working through the Traffic Impact Study 2.0 – It’s time our studies became more defensible. Stay tuned…
Mike’s Take – When we review Traffic Impact Studies prepared by other consultants, we don’t pick apart every assumption they make in the study. I think it’s unprofessional to nitpick assumptions when it is one engineer’s opinion vs. another’s. Instead, we focus on intersections or corridors where the analyses indicate they’re on the edge of needing to be upgraded. In those situations, we’ll work through the assumptions to see if tweaking them will lead to a different conclusion/recommendation. And like Bryant mentioned, I think it’s time we tighten up our process and assumptions.
Want to review our installments of the Traffic Impact Study Improvements series? Here are the links to the other articles:
Intro – Top 6 Ways to Discredit a Traffic Study
Part 2 – Would Multiple Results Help Us?
Part 3 – All Trips are Equal, But Some Trips are More Equal Than Others
Part 4 – 8 Considerations When Generating Trip Distribution Patterns
Part 5 – When is a Trip Not a Trip? How to Determine Trip Generation Types
Part 6 – Should Safety Review be Included in a TIS?
Very constructive, and professionally done post you guys. Reflects best traits of reflective professional practice.
Many ITE trip generation data sets have so much scatter that you may as well through a dart. Multi-variable equations are needed rather than the single variable used, like the movement from crash reduction factors to safety performance functions.
What would you think of using WalkScore or something like it to determine whether actual generation will be above or below average? Would you suggest any other variables, such as v/c of the adjoining street?
I’m asking because I recently reviewed a preliminary plan for an aprtment complex. Given that it’s in a car-dependent location (at least a mile from pretty much anything), I suggested that trip generation is likely to be above the ITE rate.
Jim – You raise a very good point about the ITE trip generation data set. Most of them do look like a scatter plot. Using WalkScore is an interesting idea to try to pick the nature of an area. The difficult point is then translating that to higher/lower than the ITE average rates. I say that because we know nothing about the data points in the ITE data set. By saying a low WalkScore means above average trip generation rates, you’e assuming the data points in the data set represent a range of location types. The average rate for an apartment could be reasonable in the scenario you give IF all of the data points for apartments happened to be collected at suburban locations with low WalkScores. I like where you’re headed, but the ITE data is too much of a black box to be able to make defensible assumptions. Mike
In preparing and reviewing traffic impact studies by others for over 30 years, the biggest area by far I have seen to tear apart traffic studies isn’t what is in them, but what is left out. Many times traffic studies don’t include enough intersections to get an accurate picture of what is going on. In the last few years I have seen several studies where there is gridlock a block away causing queue spillback for half a mile or more, but the studies ignored it. The counts didn’t mention that when the signal turned green the traffic couldn’t move, so the counts were artificially lower than the desired travel volumes. The studies also didn’t mention the damage done by the additional site traffic on the already massively overcapacity intersections just a block or two away. As a remonstrator’s traffic engineer, I have found the most effective way to deal with this is to take a video of existing conditions into the public hearing and present it. The planning commission and public may not understand all of the technical aspects of a traffic study, but everyone can relate to seeing a green light and having no where to go.
It is amazing how the broad range of transport related issues are so similar in North Ireland as they are in the U.S.
The main recent changes we have are use of an online planning process that makes everyone an expert in undertaking transport assessments. This can create a reactive process and politicizes the transport assessment. Local authorities may want impacts assessed for example but don’t know why i.e. engineering judgement.
I’m confused, when a TIA stated that a development will generate 1000 vehicle trips. Then it says 500 vehicles will enter the development via X street and 500 vehicles will exit the development via Y street. I am wondering what happen to the other 500 vehicles? are these other 500 vehicles staying in the development or magically exit the development in thin air? Is the Traffic Engineer who prepare this TIA not able to do maths, or is this the way of Traffic Engineering?
500 entering vehicles + 500 exiting vehicles = 1,000 vehicle trips