Travel demand models have traditionally been used to produce traffic forecasts to support long range planning applications, particularly for new facilities or those for which a significant increase in capacity is proposed. Forecasts are rarely done for near-term highway system changes that will only persist for several months. One such application, however, is to forecast peak hour traffic volume changes due to freeway lane closures during highway reconstruction. The Delaware Valley Regional Planning Commission (DVRPC) is currently preparing such forecasts to assist the Pennsylvania Department of Transportation during the reconstruction of I-95 through the City of Philadelphia. This is a multi-year construction effort that will rebuild the interstate and reconfigure several interchanges. There are numerous construction phases, with over-lapping lane and ramp closures. Each ramp closure will have one or more designated detour routes. Traffic forecasts are being prepared that will quantify the traffic volume changes along the detour routes and the impacts to adjacent facilities. These forecasts are then used to develop appropriate mitigation strategies, such as temporary traffic signals, parking restrictions, and the construction of temporary detour roads.

To prepare these forecasts, DVRPC’s regional travel demand model was used in conjunction with a detailed subarea model focused on the corridor. The model was calibrated against AM and PM peak hour turning movement counts at 85 signalized intersections and along the interstate and each of its interchange ramps. A multi-step matrix correction procedure was employed to match target entering and exiting volumes at the subarea network’s cordons, link volumes along the interstate and arterial street network, and finally intersection turning volumes. For each construction phase, the regional travel model was used to determine corridor-wide traffic volume changes and diversion from highway to transit due to interstate lane and ramp closures. The subarea model was then used to forecast traffic volumes along the detour routes and adjacent intersections.