Detection of Safe Operating Regions: A Novel Dynamic Process Simulator Based Predictive Alarm Management Approach Tama ´s Varga,* Ferenc Szeifert, and Ja ´nos Abonyi Department of Process Engineering, UniVersity of Pannonia, P.O. Box 158, Veszpre ´m 8201, Hungary The operation of complex production processes is one of the most important research and development problems in process engineering. A safety instrumented system (SIS) performs specified functions to achieve or maintain a safe state of the process when unacceptable or dangerous process conditions are detected. A logic solver is required to receive the sensor input signal(s), to make appropriate decisions based on the nature of the signal(s), and to change its outputs according to user-defined logic. The change of the logic solver output(s) results in the final element(s) taking action on the process (e.g., closing a valve) to bring it back to a safe state. Alarm management is a powerful tool to support the operators’ work to control the process in safe operating regions and to detect process malfunctions. Predictive alarm management (PAM) systems should be able not only to detect a dangerous situation early enough, but also to give advice to process operators which safety action (or safety element(s)) must be applied. The aim of this paper is to develop a novel methodology to support the operators how to make necessary adjustments in operating variables at the proper time. The essential of the proposed methodology is the simulation of the effect of safety elements over a prediction horizon. Since different manipulations have different time demand to avoid the evolution of the unsafe situation (safety time), the process operators should know which safety action(s) should be taken at a given time. For this purpose a method for model based predictive stability analysis has been worked out based on Lyapunov’s stability analysis of simulated state trajectories. The proposed algorithm can be applied to explore the stable and unstable operating regimes of a process (set of safe states), information that can be used for PAM. The developed methodology has been applied to two industrial benchmark problems related to the thermal runaway. 1. Introduction In the rush to exploit the advantages in computer-based automation, many companies in the process industry have overlooked one of the most important elements in their business value chainsthe plant operator. The plant operatorsthe forgot- ten knowledge workersis on the frontline of real-time opera- tions, making decisions that directly impact plant safety, reliability, profitability, and ultimately shareholder value. Opera- tors like other knowledge workers analyze information, diagnose situations, predict outcomes, and take action to deliver value. Optimal operating conditions of production processes are getting closer to physical constraints. Therefore, the development of knowledge based expert systems is more and more important to support operators for supporting operators to keep operation conditions in this narrow range. Beside this requirement, it is necessary that an expert system is able to detect failures, discover the sources of faults, and forecast the false operations to prevent from development of production breakdowns. 1-3 A process alarm is a mechanism to inform the operator about the development of an abnormal process condition for which an operator action is required. The operator is alerted in order to prevent or mitigate process abnormalities and equipment malfunctions. A poorly functioning alarm system is often noted as a contributing factor to the seriousness of upsets, incidents, and major accidents. Significant alarm system improvement is needed in most industries that utilizes computer based distributed control systems; it is a massively common and serious problem. Most companies have realized that they need to thoroughly investigate and understand their alarm system performance. The safe state is a state of the process operation where the hazardous event cannot occur. The set of safe states defines safe operating regions. A logic solver is required to receive the sensor input signal(s), to make appropriate decisions based on the nature of the signal(s), and to change its outputs according to user-defined logic. Beside the change of the logic solver output(s) result(s) in the final element(s) taking action on the process (e.g., closing a valve) to bring (back) it to a safe state. Defense against a possible abnormal situation can have one or more independent protection layers. To analyze the number of protection layers the method called layer of protection analysis (LOPA) can be applied. 6 The possible protection layers have a hierarchy in which the design of the process and the basic process control systems is the basic levels at the bottom. Hence, the design of a very reliable and controllable process is crucial. Hence, the development of tools to support the design of a reliable process is a really important topic. Alarm management can be applied to extract the necessary information about the crucial parts of a process for designing or improving the safety system of the process. A proper alarm management results in improved safety, reliability, and over- profitability of the process. 4 Alarm management is a fast growing, high profile topic in the process industry. The BP Upstream Technology Group proposed a five-level scale using the following nomenclature: overloaded, reactive, stable, robust, and predictive levels. 5 Technology that can achieve the predic- tive performance level is still experimental and “bleeding edge”. In ideal case, predictive performance will involve the following kinds of techniques: • Early fault detection: Early detection of process deviation from its normal operation or breakdown of a process device by monitoring a set of process variables. Deviations can be detected even when each process variable is within their operating and alarm limits. * To whom correspondence should be addressed. Tel.: +36 88 624447. Fax: +36 88 624171. E-mail: vargat@fmt.uni-pannon.hu. Ind. Eng. Chem. Res. XXXX, xxx, 000 A 10.1021/ie9005222  XXXX American Chemical Society