Optimal Reliability Design of Process Systems

Inoue, Koichi; Gandhi, Satish L.; Henley, Ernest J.

doi:10.1109/tr.1974.5215700

Cited by 10 publications

(3 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The earliest paper on the application of reliability principles to chemical engineering systems appears to be that of Rudd (1962). Since then there have been a number of others; taken in approximate chronological order these are: Fan et al (1967), A Reliability Optimization, study of a desalination unit Buffham et al (1971), A Tutorial Review of Process Applications McFatter ( 1972), Reliability Experiences in a Large Refinery provement Powers and Tompkins ( 1973), Fault Tree Synthesis Allen and Pearson (1973), Optimum Reliability Investment Gandhi, Inoue, and Henley (1973), Optimization of Process Reliability Gandhi and Henley (1974), Computer Aided Design for Optimum Systems Availability Inoue, Gandhi and Henley ( 1974), Optimal Reliability Design of Process Systems Rosen and Henley ( 1974), Optimization of Intermediate Storage Wood et al (1974), Process Plant Reliability Optimization by Linear Programming Freeman and Gaddy (1974), Quantitative Overdesign to Availability Targets…”

Section: Conclusion and Significancementioning

confidence: 99%

“…The sensitivity of the MTBF to the reliability rij of the jth unit belonging to the ith module and to the number of units in the ith module is given by and Therefore, by the addition of an integration routine to the reliability calculation program described earlier, the system MTBF and its sensitivities can be easily calculated. A computer program containing a numerical integration routine based on Simpson's formula has been developed by Gandhi and Henley (1974). This MTBF calculation routine is more flexible and easier to apply to more complicated systems than the RELCOMP routine of Fleming (1971) which is only applicable to series-parallel configurations.…”

Section: Extension To System Mtbf Calculationmentioning

confidence: 99%

“…Equation (41) is then substituted into a profit objective function such as Profit = net incomecapital costmaintenance cost the net income being directly related to the plant availability. Gandhi and Henley (1974) have published studies of availability optimizations, a$ have Allen and Pearson ( 1974). two fields have developed independently: fault tree people talk in terms of minimal cut sets; reliability graph people in teims of minimal paths.…”

Section: (41)mentioning

confidence: 99%

See 2 more Smart Citations

Process reliability analysis

Henley

Gandhi

1975

AIChE Journal

Self Cite

View full text Add to dashboard Cite

This paper presents a unified approach for obtaining system reliability and availability parameters from block diagram representations of process systems. It starts with computer algorithms for obtaining the minimal paths in a system, then shows how this information can be combined with unit reliability data to obtain the overall system reliability, and the response of system reliability to changes in unit reliability (sensitivities). It then demonstrates how these sensitivities can be used to optimize the system reliability and availability by equipment redundancy. Extension of the methodology to the automated production of system fault trees as well as the mathematical relationship between fault trees and block diagrams are noted. ERNEST SCOPEThe trend to larger, more complex, and highly interdependent process systems has been in evidence for some time. With increasing scale, repair becomes more costly in both time and money, startup and shutdown losses increase, and the vulnerability of the plant becomes a matter of grave concern. Further factors spurring the process industries' interest in reliability analysis are the high cost of maintenance, which now is as much as 10% of plant cost/annum, and the problems associated with insurance, siting, environmental degradation due to failures, and the establishment of a good public image.In this paper we focus attention on analytical methods whereby system reliability parameters can be obtained from block-diagrammatic representation of process flow sheets. Although other techniques, such as state enumeration methods are treated, primary emphasis is on the development of path enumeration methods by which the reliability parameters of complex (nonseries-parallel) systems can be computed. As demonstrated in the paper, this approach is particularly powerful because it also permits calculation of the sensitivity of the overall system reliability to changes in configuration and unit reliability. These sensitivities, being gradients, are then used in conjunction with integer, gradient-based, optimization algorithms to solve problems involving allocation of equipment and maintenance intervals.Also discussed are the construction of fault trees and some potentially useful algorithms for their automated construction from block diagrams. CONCLUSION AND SIGNIFICANCEThe increasing interest of the chemical industry in reliability analysis is clearly demonstrated by the fact that only three articles on this subject appeared in the chemical engineering literature prior to 1971, and more than 16 have been published since. To date, the process industry has depended almost completely on computational methods developed by the military, electronic, aerospace, and nuclear industries where primary concern is catastrophic failure of essentially nonrepairable systems.In attempting to apply previously developed methods to process systems, one finds that they do not readily lend themselves to many of the problems of interest to chemical engineers who generally must deal with multicomponent stre...

show abstract