The California Energy Commission is required to regularly forecast statewide transportation fuel demand under alternative policy scenarios and, in particular, assess alternative fuel scenarios in collaboration with the California Air Resources Board. To do this, the Energy Commission relies on a modeling system called DynaSim. The recently collected California Household Travel Survey data offer an opportunity for updating models. In addition, other agencies have models with similar capabilities (for example, the California Statewide Travel Demand Model). This report is the outcome of a project to review and evaluate current models and the new data sets, develop alternative options for improved models, and make recommendations. The scope is limited to fuel usage from personal vehicles in California. The review concludes that there are major flaws with the urban and intercity travel models of DynaSim, and that the Energy Commission should develop alternative modeling procedures that incorporate in some fashion the availability of the California Statewide Travel Demand Model. Two primary issues (and how they interrelate) are the quality of total vehicle miles traveled modeling and forecasting methods, and the proper allocation of vehicle miles traveled to vehicle types in the personal vehicle fleet. Both are critically important for producing fuel demand forecasts, particularly in cases involving future alternative fuels and new vehicle technologies. The report provides guidance on a future pathway of new model development with sequential testing and evaluation.

Key words: