The paper deals with risk assessment of an automated driving system considering a scenario-based testing approach, which provides evidence of the performance, reliability and safety, ensuring that the new technology is rigorously tested and refined, before deployment. The following questions are addressed: how to quantify the risk associated to a given test scenario, what is the severity of the associated hazard and what is the probability of occurrence of these scenarios? Test scenarios are generated systematically using orthogonal arrays, and depending on the strength of the orthogonal arrays, a statistical indication of test coverage is available. Furthermore, scenario complexity is quantified using Shannon information content. The proposed risk assessment methodology, is aligned with the existing standards, is straightforward, and is illustrated via an example.