Résumé : We calibrate a SEIR contagion framework, with an extended transmission rate, Rt that account for two different classes of infections, and for the effect of both risk perception on citizen behaviors as well as government actions on Belgium date. We demonstrate that curbing the Covid 19 pandemic and g enough hospital care capacity structurally with untargeted social distancing is challenging, requiring to sustain at least a recurrent cut on 50% of contagious contacts on top of creating unknown economic risks. With sufficient specificity (>70%) of economy. In practice, however, testing the full population is not possible, and would need to randomized requiring for safety a mix of testing and tracing, together with some other measures, like protection of more susceptible population either by nature (elderly), or by economic necessity (e.g. front lines workers).