Fault diagnosis requires the knowledge models which describe the behavior of the chemical processes. However it is cumbersome and time consuming work to generate these models. Also the knowledge acquisition required for generation is difficult. The objective of this study is to examine whether the knowledge aquisition for fault diagnosis knowledge models, especially bottom-up models, is possible through qualitative simulation. Qualitative simulation is based on the study of modeling, representation of feedback control and state transition using the QSIM algorithm. Also in this study, qualitative simulation is applied to a buffer tank level control system using the simulation strategy of this paper. The results of the simulation show several behaviors of the processes and the usability for the generation of a knowledge model. However when several different results of simulation are generated, some are spurious solutions which result from insufficient information.