A company which has a production facility has a strong concern for early business recovery after a natural disaster. Though the "Recovery Time Objective" (RTO) is very important to carry out business continuity management, a method to predict restoration time which is necessary to decide RTO has not been established. This paper proposes two simplified methods applying the idea of an event tree analysis for productive facilities. However event tree calculation becomes difficult when the number of events is larger. Commonly used PC programs cannot do the calculation which involves over thirty events. The event tree analysis for over thirty events is able to be calculated by the restoration time evaluation of a series of scenario simplified. This paper shows the derivation of the methods, the calculation accuracy and the validity of the methods. Finally a calculation example to which these proposing methods are applied to an assumption factory is introduced. It is found that recovery time does not become short though the number of restoration staffs is larger because of influence of other factors like approval time and delivery time.