In this paper, we investigate a new type of retrial queueing model with feedback and working breakdown services. The regular busy server may become defective by disasters (negative customers) at any point of time. Negative customers arrive only at the service time of a positive customer and remove the positive customer from the service. At the instant of failure, the main server is sent for repair and the repair period begins immediately. During the repair period, the server gives service at a lower speed (called working breakdown period). The steady state probability generating function for system size and orbit size are obtained using the method of supplementary variable. We also obtain some analytic expressions for various performance measures such as system state probabilities, mean orbit size, and mean system size of this model and some important special cases are discussed. Finally, some numerical examples are presented to study the impact of the system parameters.