Traffic Impact Study Process – Part 5: Prepare Traffic Forecasts

Mike Spack and Bryant Ficek have literally written a book about Traffic Impact Studies and the process from authorization to final study. We present the information we wish we had when starting our careers and hopefully have tips and refreshers that even experienced traffic engineers will find useful. This series presents the basic steps of completing a Traffic Impact Study from their book. See the early posts here.

Our previous parts 1 through 4, collecting the data/background information and laying out your assumptions, are like chopping the vegetables and measuring out the spices when you’re making dinner. Now, it’s time to start cooking. We are now onto first engineering step in the traffic study process – preparing the traffic forecasts.

Assuming you don’t have a reliable crystal ball, and why are you reading this if you do, you’ll want to follow the industry standard methods for preparing traffic forecasts. The goal is to develop credible traffic projections to use in your capacity analyses. You’ll implement this same process whether forecasting daily traffic volumes or peak hour turning movement volumes for your traffic study.

The basic procedure involves taking your existing traffic volumes and increasing them in some manner to account for a future condition that doesn’t yet have your development. This is typically called the No-Build or Without Development scenario. Then you add on the traffic from your specific development, and you get what is usually referred to as the Build or With Development scenario.

Calculating No-Build Traffic Volumes

Traffic on most roads that have more than a few hundred cars follows the standard s-curve shown in the figure below. If you memorized your stats textbook, you’d recall this is also called a logistic curve or cumulative distribution function. Traffic volumes on roads are quite stable for a certain period of time after the road opens. They experience an exponential growth phase and later have a period of linear growth. Finally, the traffic volumes plateau through a logarithmic phase. If you’re in a fully-developed area of town, it’s relatively safe to assume that traffic volume growth is well into the logarithmic phase and you can employ a very small (0.5% per year or less) linear growth rate to estimate future traffic growth. If the study area is in a developing part of town, however, forecasting traffic is much trickier because the traffic volume growth is likely in the steep part of the s-curve in either the exponential or linear phases.

Traffic Growth on a Typical Road

Even though the old axiom holds true that past performance is no guarantee for predicting future results, traffic engineers do like to assume linear growth. They do this mainly because it’s the simplest way to develop future traffic volumes, while also being defendable. Here are five methods to develop your growth factor:

  1. Use the area’s blanket growth factor. For instance, the Minnesota Department of Transportation (MnDOT) develops a growth factor for each county in the state. If the 20-year factor for your County is 1.2, that’s straight-line growth of 2% per year.
  2. Use historical daily traffic volumes. Ideally, you would have five or more data points or years of traffic volumes. Then you can estimate the data and pick out a growth rate to the nearest 0.5% or average the growth rates.
  3. Use another recent traffic study completed within your study area. The hard work has already been completed for you and has hopefully been approved by the reviewing agency. You can use the same rate and cite this study as the source. Don’t forget to account for that study’s development as part of your background traffic.
  4. Use the agency’s transportation plan. Most areas have a comprehensive or individual transportation plan that may include 20-year forecasts. You can calculate a growth rate from those forecasts in your study area. Be wary of over-estimating if you have a large development. Some or all your development traffic may already be accounted for in the agency’s forecasts.
  5. Use the agency’s regional transportation model and develop your own If you’re lucky, you’ll find a very nice person at the agency responsible for the model that will do this for you. This method provides the truest, or at least the most defendable, no-build traffic forecast condition. Using the regional model is often quite cumbersome (and expensive) and only justified on very large projects, such as a regional shopping mall or a 1,000-acre subdivision.

No matter which method you use, the goal is to create a defendable linear growth factor that can be applied to your existing traffic volumes to calculate a future year No-Build scenario. Although it could be argued that growth rates for different roads may be more logical, it isn’t feasible to develop peak hour turning movement counts going between intersections based on multiple growth rates because they simply wouldn’t balance. You’ll need to choose one of these methods, or a combination, to develop a linear growth rate you believe will be reasonable for your project area.

Also, if there is a nearby development that was recently approved but not yet built, it may be prudent to explicitly account for traffic from that development in addition to the no-build forecasts you developed based on the linear growth rate. You’d be slightly double-counting, but it’s certainly worth considering.

Once you’ve determined a growth rate, apply that to your existing daily volumes and existing peak hour turning movement volumes to develop the future year No-Build traffic forecasts.

Calculating Traffic Volumes Due to Your Development

There are two elements to forecasting how much traffic will enter and exit your development:

  1. Determining the quantity of the traffic, and
  2. Determining the route that traffic will travel.

Trip Generation

The industry standard for determining how much traffic will be generated by development is the Institute of Transportation Engineers’ Trip Generation Manual. (Jargon Alert – a trip end is one car entering OR one car leaving a development. A single car that goes to your development and subsequently leaves the site adds up to two trips.)

We will also point out our effects to collect local trip generation data, available at TripGen.org. We believe, and ITE supports, the notion that local data reflects the areas driving habits and tendencies better than a national database. Other local jurisdictions and businesses sometimes have their own trip generation as well. For instance, Las Vegas, Nevada has trip generation for casinos along the Las Vegas Strip and LifeTime Fitness commissioned various studies for trip generation data associated with their operations. Additionally, there are national datasets around the world, such as the TRICS dataset that covers Ireland and the United Kingdom.

Another option is to collect trip generation at three similar land uses as part of your study. Data collection is less expensive using portable video camera systems and fast forward viewing – such as the CountingCars.com system. This method produces the most accurate and up-to-date rates for your project.

Whatever source you use, the basic procedure is to multiply the average trip generation rate for your land use and study period by the units in your development, and then apply the distribution split. For example – single-family homes have an average trip rate for the weekday a.m. peak hour of 0.74 vehicles per dwelling unit with 25% entering and 75% exiting. A subdivision with 200 homes would therefore generate 148 vehicles in the a.m. peak hour – 37 entering and 111 exiting vehicles.

Deductions to Trip Generation

It’s relatively uncomplicated to calculate the traffic coming in and out of a single-family home subdivision in the suburbs. These are new houses and the residents will predictably be driving to work, the grocery store, and to school to drop off the kids. However, there are several scenarios where a new development doesn’t actually increase the total number of cars on the road, but shifts or redirects traffic. The three basic types of deductions to apply to the raw trip generation are Pass-By, Diverted, and Internal trips.

Pass-By trips are the most typical deduction and reflect drivers already on the road who include a stop at a new commercial/retail development in the future. A typical gas station demonstrates the difference between New and Pass-By trips. A “new trip” to a gas station is a customer who runs out of gas for the lawnmower, jumps into the car, goes to fill up the gas can and then goes back home to use the lawn mower. A “pass-by” trip is a customer who’s driving home from work, realizes he needs gas in the car and pulls in/out of a convenient gas station on the way. He’s already on the road in this situation and stops in as a “pass-by.” In your forecasts, you add the “pass-by” vehicle to the turns in and out of the driveway, but you subtract the vehicle from the through moving traffic. Thus, you have a reduction in through traffic to offset the turns in/out right at the driveway, but traffic doesn’t change at the intersections upstream or downstream from the traffic.

Diverted trips are related to Pass-By, but reflect the using one or more side roads to reach the development. Consider a person is driving home and needs to stop to grab takeout for dinner. His favorite takeout restaurant is two blocks from the main road he uses. He then “diverts” off his normal path to get the takeout and then goes back to his route to continue home.

The trouble with identifying diverted trips is that the deduction depends on how far out your study area extends. Think about stopping in at Home Depot on the way home from work to grab a box of nails. The driver must exit the freeway and drive half a mile on a county road to get there. When she’s done, she goes back down the county road and gets back on the freeway. If we’re only analyzing the intersection on the corner of the Home Depot, looks like a new trip because she’s coming and going in the same manner through the study intersection. Now, if you are also analyzing the two ramp intersections at the freeway interchange, you must treat it as a diverted trip. She’s coming and going from the same direction at the intersection on the corner of the Home Depot, BUT she is also continuing onto the freeway in the same direction from where she started. She is not returning to the place where she originally got into his car. Thankfully, diverted trips usually only come into play in traffic studies for large, regional shopping destinations.

Internal trips are the third type of deduction you may need to account for in your trip generation and represents interactions between multiple buildings within the proposed development. If your development has a big-box retail store with a bank and fast food restaurant on the corner of the site, some of the people who go to the fast food restaurant will also be coming from or going to the big-box retail store and may include a stop at the bank. That’s three entering trips for the development, but only one impacts the surrounding roads.

Several methods are available for calculating trip reductions, such as the ITE Trip Generation Handbook, NCHRP Report 684 – Enhancing Internal Trip Capture Estimation for Mixed-Use Developments, or the Mixed-Use Development (MXD) Trip Generation model developed for the US EPA. We recommend you consider using one of these tools on mixed-use projects, but you should have a good amount of experience before you’re preparing a traffic study for a project complicated enough to justify using the methodology.

Beyond those basic reductions, additional trips could be removed based on the location of the site and the implementation of a Travel Demand Management Plan – sometimes called a TDMP or TDM Plan. The ITE trip generation may not fully account for pedestrian, bicycle, and transit trips. A development site in downtown or adjacent to a busy transit line, like a permanent light rail station, will have less traffic than the national averages. It’s justifiable to reduce the trip generation in those sites to account for people walking or riding the bus. Before taking this reduction, be sure to consider other factors like the development design, the amount of parking and whether it is free, proximity to the bus or train stop, etc.

Similarly, a TDMP details strategies to encourage travel outside the peak hours and by methods other than a single-occupancy passenger car. Typically a binding agreement between the City and the development, examples of encouragement are subsidizing bus passes, installing bike racks, providing carpool matching and staggering work schedules. With an active TDMP, even a suburban development in a sea of free parking will reduce the trip generation to a degree.

Trip Distribution

Once you have the base trip generation calculated (and sub-divided into new, pass-by, diverted, and internal if needed), distribute those trips on your study roads for daily traffic volumes and through your study intersections for peak hour analyses. The distribution pattern means determining which road(s) each trip will use to and from the proposed development. This process is usually the least scientific part of the traffic forecasting process.

The basic methods to determine the trip distribution for your site include:

  • Using the daily volumes on the study roads access points to and from your site. Sum the volumes and divide each daily volume by that total to determine a percentage for that road. For example, if the three ways into your study area have daily traffic volumes of 32,000, 16,000and 45,000 vehicles per day; round the to/from distribution pattern to 35%, 15%and 50%, respectively.
  • Consulting area maps to review access to regional roads (interchanges), population centers, and nearest competing land uses are. For instance, if your study is for a new WalmartTM, figure out where the other area WalmartTM and TargetTM stores are. If there are stores to the west and south, your distribution pattern should probably focus toward the north and east.
  • Use the existing distribution around the site based on turning movement counts. This is especially true if your development is an addition to an existing If you’re adding a wing to a hospital, base your distribution pattern on the existing driveway turning movement counts.
  • If your project is retail-oriented, you may be able to get a trade area/demographic study from the developer that provides a distribution pattern. Our experience suggests these studies are tightly guarded and not shared with traffic engineers very often, but it’s worth asking.
  • If your project involves a school or club, you may be able to get zip codes for all their students or members. This is a great way to develop the distribution patterns for those types of uses and these types of developers are much more willing to share this information than the retail folks.
  • If you can get into the regional model, you may be able to extract distribution pattern percentages for your study parcel, but this is rarely done. Count yourself lucky if you have a friend at the local Metropolitan Planning Organization who will pull out origin-destination tables for you to base your trip distribution pattern.

After establishing the general percentages to the surrounding edges of your network, determine the specific paths to and from your development’s access driveways. Most drivers will take the most direct path and incoming/outgoing routes are likely the same. However, one-way roads, limited access intersections, and other factors may result in several different routes.

Once you’ve determined your trip generation numbers and the trip distribution pattern, developing the volume due to development portion of the forecasts is elementary math. You can either do this mathematical process by hand, but many analysis software programs will complete the work for you.

Calculating the Build Traffic Forecasts

The Build or With Development forecasts are also relatively basic to determine – Add the volume due to development to the No-Build volumes and, presto, you have your Build condition forecasts. Be sure to keep your scenarios and study periods in order so you are adding the proper development traffic to the appropriate matching No-Build traffic numbers.

For all scenarios, it’s OK to round your forecasts to represent an approximate level of accuracy. Recall an academic discussion on significant digits: just because a calculator comes up with a number of eight decimal places doesn’t mean the real number is that refined. For daily traffic volumes, round them to the nearest 100 vehicles. During the peak periods, round each turning movement volume to the nearest ten, unless it’s a low-volume movement that has a couple of vehicles, then round them to the nearest five.

If you’re using a software package that does these calculations for you, it’s possible there isn’t a system to round your forecasts. In that situation, be certain to discuss the level of accuracy in your traffic study report.

Remember to do a quality control check on all your forecasts. Ensure the numbers add up correctly. Double check that No-Build combined with the development volume does indeed equal the Build. Make certain development traffic volumes add up appropriately between your study intersections; cars generally don’t disappear. Volumes may not match between intersections if there’s a reasonable explanation, like on-street parking or driveways, but be certain that any discrepancies are truly reasonable. 300 cars parking on a street block during your peak hour isn’t reasonable. Finally, check to see if your graphics of the vehicles entering and exiting your development match the numbers on your trip generation table.

Although that seems like a lot of extra checks and balances, keep in mind that on controversial projects opponents of the project will be doing the math. They will happily point out those discrepancies in an attempt to discredit your study. Try not to give them that opportunity.

This is a multipart series. Check out the other articles in the series.

Did you like this discussion regarding Traffic Impact Studies?

Our Traffic Studies Manual contains more details, checklists, and real-life situations Mike and Bryant have dealt with on their projects. Also check out our Traffic Impact Study Report Template. This is the template we use when creating reports and is completely editable in MS Word.

At Spack Academy you can also find other guides, case studies, research briefs, and spreadsheet tools that help us be better traffic engineers. Visit us at www.SpackAcademy.com, connect with us on LinkedIn, or send us an email with your thoughts and questions.

 

Bryant Ficke Bio

 

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 *

One thought on “Traffic Impact Study Process – Part 5: Prepare Traffic Forecasts

  1. Spack Consulting,

    Looking for your professional opinion and insight with regards to trip generation of warehousing..

    Let’s say I’m proposing to build 3 large warehouses in close proximity that are all 1.0M SF each. I intend to own the warehouses and find tenants for them. I am anticipating having 3 or even 4 tenants occupying the warehouses (1 of the warehouses could potentially be split internally and occupied by 2 tenants).

    My question is, what is the most appropriate way to forecast the trip generation for these warehouses? The answer, I’m sure, will ultimately depend on how the data in the ITE Trip Generation manuals was collected in such instances. Is it appropriate to generate the trips based upon one calculation of 3.0M SF of warehousing? Or should the trips be forecasted using 3 separate calculations of 1.0M SF of warehousing? The different methodologies can result in generation variations in the hundreds of trips during the peak hours (according to 9th Edition).

    Your insight on the matter would be greatly appreciated.

    Regards,
    Chuck