The importance of dissolved oxygen (DO) concentration control in aeration tanks of a bioreactor at flow-through wastewater treatment plant (WWTP) can easily be justified by technological requirements as well as simple economics. Firstly, appropriate levels of DO concentration are essential for the vitality of microorganisms that comprise the bioreactor. Secondly, the costs of DO concentration control related to the blower station operation constitute up to 75% of the plant-wide electric bill. This paper addresses a problem of Takagi-Sugeno fuzzy model identification of DO dynamics for control design purposes. In the proposed approach the fuzzy partitions and the corresponding fuzzy sets are identified based on the knowledge of the process. Local model structure is obtained based on a larger well-established cognitive model from which a well-defined parameter set follows. Thereafter, a parameter identification is performed. A case study example of a bioreactor at Kartuzy WWTP illustrates the workflow of the proposed identification algorithm.