Process alarm management can be approached as a pattern recognition problem in which temporal patterns are used to characterize different typical situations, particularly at startup and shutdown stages. This paper focuses on learning the temporal patterns, in the form of chronicles, by extending the previously proposed Heuristic Chronicle Discovery Algorithm Modified HCDAM. The proposed extension incorporates knowledge, in particular in the form of so called temporal runs, to focus the learning process and produce less conservative chronicles. The resulting Chronicle Based Alarm Management (CBAM) approach is hence based on a diagnosis process which permits situation recognition and provides the operators with relevant information about the failures inducing alarms flows in the startup and shutdown stages. The event sequences that represent a process situation are generated by simulation and including temporal runs, the chronicles are extracted using the extended version of HCDAM. Finally, the conclusion and future work are presented.