In recent years, more and more regions start to conduct university student travel surveys to understand their travel behaviors. Many survey results indicate that university students in general have higher trip rate, shorter trip distance, different time of day distribution, and higher shares of non-motorized and transit trips compared to the general population. However, the differences are not well represented in travel demand models. The typical practice is to treat university students the same as the general population. Some models add the trip purpose of Home Based University (HBU) to model the unique trips made by university students. But they still assume university students and the general population are the same for other trip purposes. These current practices might be an issue for some regions. For example, in the Triangle region in North Carolina, university students only account for 5% of the population, but they make 47% of the transit trips. If university students are modeled together with the general population, the mode choice model would overstate the propensity for the general population to make transit trips. Therefore, modeling university student trips separately from the general population will greatly improve the performance of travel demand models in some regions.

This paper presents the approach to model university student trips separately in the Triangle region. The challenge is to separate university students living off campus (off-campus students) from the general population. Off-campus students are counted as household population in census, so they are included in the zonal population. Although some universities could provide off-campus students’ home addresses, the data quality is usually poor. It is also an issue how they can be consistent with the population synthesis process, which disaggregates the zonal population (including off-campus students) into groups of people with different characteristics.

The Triangle Regional Model developed an off-campus student residence location model to estimate the zonal number of off-campus students based on a university student survey. The off-campus students, as well as the students living on-campus, are then modeled explicitly based on the survey data to better represent university students’ unique travel characteristics.