In order to meet the requirements of locating fault causes of continuous work equipments or units in petrochemical enterprise quickly, an ontology based fault tree construction method was proposed. In the proposed method, the fault tree is generated from domain ontology, and then the ratio of efficiency to time for fault diagnosis is obtained through the quantitative analysis of the fault tree. At last, the optimal fault diagnosis path can be obtained according to the descending order of the ratio of efficiency to time for fault diagnosis. This method combines the advantages of ontology and fault tree, and it realizes rapid fault causes locating based on knowledge sharing and reuse. Application example shows that the efficiency of fault diagnosis can be increased and the costs of enterprise's maintenance can be reduced by using the proposed method.