A system dynamics-based model for demand forecasting in PPP infrastructure projects ? a case of toll roads

The success of Public Private Partnership (PPP) infrastructure projects is highly dependent on the demand for the services provided by these projects. The demand forecasting process is complex because of the influence of various economic, social and technical factors and the interrelationships among them. In addition, this process is dynamic in nature as many of these factors are time dependent. Current models used for demand forecasting have failed to account for many of these aspects. Among various modeling techniques, System Dynamics (SD) is a promising method for modeling systems with complexity and dynamicity features. The modeling process using SD can be broadly divided into Qualitative System Dynamics and Quantitative System Dynamics. This paper describes the development stages of a conceptual Qualitative SD model for demand forecasting which include: factors identification, creating Causal Loop Diagrams (CLDs), and the CLDs validation. As expert knowledge and perceptions are key requirements to develop a realistic SD model, the paper will emphasis on the knowledge elicitation involved in the development stages. The paper articulates different approaches used to collect and analyze perceptions solicited from experts in toll road projects and the demand forecasting discipline in order to build this qualitative model. In addition, it depicts how the information has been integrated into the different stages of the modeling process. The developed qualitative model will form the basis for the development of the quantitative SD model aiming at improving the practices of demand forecasting in PPP toll road projects.