Peak spreading is relevant in several types of analyses, particularly analysis for capital construction investments, air quality analysis for conformity requirements, and analysis for transportation demand management investments. This review was conducted in response to issues raised at the Washington State Department of Transportation (WSDOT) regarding the benefit/cost assumptions and calculations that could or should be made regarding the phenomenon of peak spreading. This report identifies the transportation planning issues associated with peak spreading, reviews efforts that have been made to account for it in analysis, and makes recommendations specific to the priorities of the State of Washington.
Four categories of analysis approaches were reviewed: (1) a post-processing technique in which hourly factors are applied to the daily traffic volumes output by a forecasting model; (2) peak spreading adjustments that were made to the four-step modeling process; (3) attempts to develop more sophisticated stand-alone peak spreading models, which could then be used as sub-models within the more traditional forecasting process; and (4) stand-alone models that were completely independent of the four-step forecasting process.
Because consistent statewide forecasting methods have not yet been implemented, peak spreading analysis methods developed for WSDOT in the short term should be independent of four-step forecasting models. However, the establishment of a common travel demand forecasting framework throughout the state would definitely make longer-term modeling approaches more feasible. In the short- to mid-term, directional historical traffic data that have been collected by WSDOT should be compiled for key freeway locations. These historical traffic profiles could be used to formulate simple models on the basis of future estimated growth rates to predict future traffic conditions. In the longer term, a departure time element should be included in the ongoing research at the University of Washington, the goal of which is to include a more robust variety of traveler choices in travel demand forecasting.