Multivariate alarm systems to recognize rare unpostulated abnormal events

Most chemical and manufacturing plants have safety/reliability systems in place thatare well equipped to handle commonly occurring postulated abnormal events, butoften prove to be ineffective in predicting highly unanticipated and randomly occur-ring unpostulated abnormal events. In this paper, an advisory system is developed toanalyze and recognize such events, consisting of novel, multivariate alarm systemsand response actions introduced using process modeling and path-sampling forunpostulated abnormal events. It augments existing safety/reliability systems, sug-gesting actions when unanticipated abnormal events are approached. Forward-fluxsampling (FFS), developed to discover rare molecular dynamics pathways, is applied.With an approximate process model, for an exothermic continuous-stirred tank reactor using a perturbed feed concentration, the FFS algorithm is applied to identify raretrajectories between high-conversion and low-conversion steady-states, with keyprocess variables saved at variouscrossing points.Then, committer probabilities,p(B), are computed at each crossing point, yielding a mathematical model thatexpressespBas a function of the key process variables; that is, the reactor tempera-ture (T), cooling-water flow rate (F-C), and cooling-water temperature (T-C), selected aschoices for the primary alarm variables. For these, alarm thresholds; that is, L-low,LL-low low, and LLL-low low low, are suggested by computing their critical ranges,given the p(B )ranges for every alarm threshold.

Automated summary

An advisory system is developed in this paper to analyze and recognize highly unanticipated and randomly occurring abnormal events in chemical and manufacturing plants. The system utilizes novel multivariate alarm systems and response actions, along with process modeling and path-sampling techniques. This system aims to enhance the existing safety/reliability systems by suggesting appropriate actions when approaching unanticipated abnormal events.