In
a continuously operated chemical plant, not only the critical
unit may fail after a long period of operation, but also its process
load may vary from time to time. Thus, it is necessary to incorporate
standby mechanisms to maintain uninterrupted production and also satisfy
the unexpected increases/decreases in external demands throughout
the operation horizon. Although a few related studies were reported
in the literature, a comprehensive analysis of multilayer standby
mechanisms in continuous processes still has not been carried out.
To address this need, a generalized mathematical programming model
has been constructed in this work to automatically synthesize the
optimal designs and maintenance policies of the standbys for any continuous
process by minimizing the total expected lifecycle expenditure. A
MATLAB code has been developed to execute the optimization runs via
genetic algorithm. The feasibility and effectiveness of the proposed
model and solution method are demonstrated in this paper with case
studies concerning the pump systems in a typical chemical plant. From
the optimization results, one can obtain proper design specifications
of the standby mechanisms, which include: (1) the number of protection
layers (or online warm standbys) and the number of cold standbys stored
offline; (2) the numbers of installed sensors in each measurement
channel, the corresponding voting-gate logic, and their spares; and
(3) the inspection interval of switch and the number of its spares.