Constructing Fault Trees for Advanced Process Control Systems—Application to Cascade Control Loops

Abstract

A systematic approach to construct fault trees for advanced process control systems is presented in this paper. For illustration purpose, the proposed method is explained with a specific feedback scheme, i.e., the cascade control strategy. The digraph configuration of a standard system is first described and analyzed in detail. On the basis of a series of qualitative simulation studies, all failure mechanisms can be identified and summarized with a set of generalized fault-tree structures. The fault trees produced with the conventional digraph-based techniques are shown to be not as comprehensive as the ones constructed with the proposed approach. To demonstrate the correctness of our analysis, the successful application of the proposed structures to a heat exchange process is presented. In addition, the resulting fault tree is compared with one obtained from a single-loop feedback control system and the trade-off between the two in system reliability and control performance is assessed accordingly.

From Model, Signal to Knowledge: A Data-Driven Perspective of Fault Detection and Diagnosis

“…FT originally developed at Bell Lab in 1960s is a logic cause-effect tree that propagates primary events (faults) from bottom to the top level events (symptoms). A recent application of FTs in FDD was reported in [57] for reliability analysis and fault diagnosis. SGD is a graph with directed arcs leading from a 'cause' node to 'effect' nodes and these arcs are given a positive or negative sign.…”

Development of an Optimal Sensor Placement Procedure Based on Fault Evolution Sequences

“…Thus, the complete fault propagation process in a NFBL is often described on a case-by-case basis. 30 However, if the NFBL in Figure 1c is a control loop, then the final steady-state values of the loop variables can be assigned according to the general guidelines suggested by Ju et al 32 Table 1 is a listing of the final states of control NFBLs in various scenarios. Thus, the FPP corresponding to D X (+1) in this situation can be expressed as It can be observed that the final states of all loop variables in this FPP are lumped into a single ending node in a square bracket and their precedence order is left unspecified.…”

Systematic Enumeration of Fuzzy Diagnosis Rules for Identifying Multiple Faults in Chemical Processes

“…Let us assume that the combined effects of m 2 (-1) and m 3 (+1) are equivalent to those caused by an uncontrollable disturbance. Consequently, the final states of loop variables can be assigned by following the approach suggested in the work of Ju et al, 24 i.e., h(+1), s 5 (+1), s 6 (-10), and m 1 (-10). The corresponding simulation results are shown in Figure 10a.…”