What the ITE Trip Generation Manual Isn’t Telling You

University-123RFThe Institute of Transportation Engineers’ (ITE) Trip Generation Manual suggests that “the user may wish to modify trip generation rates presented in this document to reflect … special characteristics of the site or surrounding area.” We translate this suggestion as a directive to obtain local data when possible because it will be more accurate than a national average. And now we have further proof of that.

Our friends from the Florida Department of Transportation, Gary Sokolow, Nathan Hicks, and Michael Stafford, provided us a wealth of new trip generation based on their work. (Thank you gentlemen!) Combined with our trip generation information, we can begin to see how these local characteristics can result in significant differences between rates. The trip generation rates for a gas station and student housing are shown in the tables below.

You can review the full data and more at www.TripGeneration.org. That’s our free site to share trip generation data with everyone, with currently over 5,000 hours of professionally collected traffic data for popular land uses.

Button - Get Trip Generation Data v2

GasStationChart

When comparing rates for the gas station, ITE underestimates both Florida and Minnesota data. The Florida rates are closer, particularly the peak hour rates. Since most traffic impact studies examine the peak hours, the difference is probably within the margin of error and not likely to change the final recommendations. Minnesota is at least 24 percent higher across the board. Imagine missing one out of every four cars in your study and you can easily see how major impacts could be missed.

Why are Minnesota rates higher? It’s likely a combination of local factors, such as cold winters that keep many people from walking or bicycling more than a few blocks and suburban areas with few transit options that force people to use their cars more often.

StudentHousingChart

Turning to Student Housing, we can clearly see Minnesota rates are significantly lower while Florida rates are significantly higher. Local factors are again at work. In Minnesota, parking is scarce and expensive. With many housing options on campus or very near campus, the short walk to class does not require a car. In Florida, multiple parking garages are available and parking is viewed as “free” since the transportation fee is combined with tuition. A lack of pedestrian/bicycle/transit infrastructure could also be a cause here.

Whatever the exact reason, the differences between the Florida, Minnesota, and ITE trip generation rates are clear. With often expensive infrastructure decisions depending upon our analyses, local trip generation is obviously the best way to account for each area’s diverse set of circumstances.

Want to partner with us to collect your own local trip generation data? We have a limited pool of COUNTcam video collection products that we’re lending for free specifically for this purpose. Contact Sales at CountingCars.com or leave your contact information in the comments section if you’re interested.

Please note: I reserve the right to delete comments that are offensive or off-topic.

Leave a Reply

Your email address will not be published. Required fields are marked *

4 thoughts on “What the ITE Trip Generation Manual Isn’t Telling You

  1. Thank you for all of your work in promoting open data solutions, such as TripGeneration.org. I, too, have been working on an approach to this problem and see opportunity with new technologies emerging. We share a philosophy on trip generation data points as I believe it should be provided as openly as possible to the transportation industry. I believe the true value will be in building applications to better share and apply the data, instead of gate keeping it. This is an area in which I am critical of the ITE.

    I am currently finalizing my research and writing a white paper to lay out my idea and hope to share it with you and the transportation community later this year.

    Cheers!

  2. Interesting results. Trip gen is just fascinating at the number of things that can affect it.

    Looks like your MN peak hour data matches pretty well with the ITE land use 853 for convenience store with gas pumps. 853’s daily is much higher though. I’ve always wondered how sites were distributed into convenience store with pumps versus gas station with market. Not very many pure gas stations around here anymore (TX).

    An interesting challenge I’ve faced recently is the giant store and gas station as a tourist attraction, the local Buc-ee’s chain being the example. 60ksf of store with 90+ gas pumps. Just a little outside the ITE gas station data points.

    Suggestion for future data collection: high-turnover sit-down restaurants. Another area where sometimes the differences between fast food, regular restaurants, and quality restaurants are fuzzy. But the main point is the AM trips for sit-down restaurants. It’s a huge number, higher than the PM, since I’m assuming it is collected at Denny’s, IHOP, and other breakfast-focused restaurants. But the *average* sit-down restaurant (TGIFridays, Chilis, etc) is usually not open in the adjacent road peak hour at all. This is also true for a smaller percentage of fast-food places that don’t open for breakfast. Rarely does anyone do trip generation observation on restaurants when they are closed, so the data is all skewed. If you don’t adjust arbitrarily, a site with a lot of restaurants will be artificially high in the AM, and probably the daily too. Now just putting data from the 70% of closed restaurants into the mix would make the rates/equations artificially low for those restaurants that are open, so instead I’d like to see that category get split into restaurants with and without breakfast service.

    • Scot – I agree. We’ll add high turnover restaurants to our list. Key with our data set is we fully describe the location in the dataset. Eventually, we’ll build a filterable system. This means you’d be able to easily get to the high turnover restaurants that serve breakfast vs. looking at a jumbled dataset and have to make interpretations. Sorry – no 90 pump gas stations around here – that’s crazy! Mike