Despite significant efforts to make operation of chemical plants safer, the occurrence of incidents clearly indicates the need for better design approaches. Studies to identify the root causes of incidents in hydrocarbon industries reveal that poor design and inadequate control systems contribute to more than 20% of the offshore incidents [1] and 30% of the thermal runaway incidents [2] analyzed. Characterizing and quantifying process safety performance is a complex problem. Traditional control engineers used the concept of phase margin and gain margin to measure the stability of single feedback loops. Although it can be viewed as a measure of safety, the method does not account for multiloop interactions and the presence of constraints in the system. More recently, researchers have used model predictive control (MPC) theory to address safety concerns. The objective of the MPC optimization problem is maximization of cost and other performance metrics, where safety is modelled as a set of additional constraints that must be enforced. The approach is not adequate as there is not a clear method to quantify the safety performance for application in design. Process safety engineering concepts emerge from cause-effect based analysis like HAZOP analysis, fault trees and event trees. These methods do not account for multivariable and non-linear interactions. The objective of this research is to develop an approach for the process control problem with safety as the primary target.In this paper, the concept of dynamic safe set (DSS) is formulated. The DSS is a set of states of the process that guarantee enforcement of safety critical constraints, in the presence of bounded safety threatening disturbances. Already existing mathematical concepts from the systems literature, namely maximal output admissible sets [3,4] and the reference governor theory [5,6] are used for evaluating the DSS. The DSS is calculated around a steady-state operating point. It is safe in the sense that if the initial state belongs to the DSS, then for all modeled disturbances the closed-loop system is guaranteed to not violate the constraints at any time in the future. The safety threatening disturbances that can increase the possibility of safety constraint violation by pushing the system to a risky operation zone are also modeled while calculating the DSS.A method to quantify the size of the DSS is also proposed by defining the concept dynamic safety margin (DSM). It is defined as the minimum distance of the steady-state operating point from the boundary of DSS. The DSM margin is relevant and important because it is not possible to model all possible disturbances. That is, a DSS with larger DSM will be able to handle unmodeled random disturbances that push the states away from the steady-state. This will be used as a safety performance metric for control system design. This will lead to designing processes with safety as the primary objective and all other performance metrics are treated as secondary considerations.The DSS approach is also extende...

No abstract

When a fault occurs in a process, it slowly propagates within the system and affects the measurements triggering a sequence of alarms in the control room. The operators are required to diagnose the cause of alarms and take necessary corrective measures. The idea of representing the alarm sequence as the fault propagation path and using the propagation path to diagnose the fault is explored. A diagnoser based on hidden Markov model is built to identify the cause of the alarm signals. The proposed approach is applied to an industrial case study: Tennessee Eastman process. The results show that the proposed approach is successful in determining the probable cause of alarms generated with high accuracy. The model was able to identify the cause accurately, even when tested with short alarm sub‐sequences. This allows for early identification of faults, providing more time to the operator to restore the system to normal operation.

No abstract