The paper proposes a method to compute the reliability of a pharmaceutical plant. A decision support system for the concerned plant was also developed. The pharmaceutical plant mainly consists of nine subsystems working in a series. One subsystem, namely, the Rotary Compression Machine was supported by standby units with perfect switch over device. The remaining eight subsystems were prone to failure. Mathematical model of pharmaceutical plant was developed using Markov birth-death process. The differential equations were developed on the basis of probabilistic approach using transition diagram which were further solved using the matrix method. A program was developed in C language to study the variation of reliability with respect to time. Steady-state availability of the concerned plant was also calculated in order to develop the decision matrices which provide various availability levels for different combinations to failure/repair rates of different subsystems. This enables the management to quantitatively analyze all the factors, viz., maintenance strategies and states of nature, which influence the maintenance decisions associated with the pharmaceutical plant. The findings of this paper would be highly beneficial to the plant management for the corrective and timely execution of proper maintenance decisions and hence enhance the performance of the system.

Increasing demand for medicines and heavy initial investment for the installation of a new plant or the expansion of the existing plant have forced the government to realize the importance of reliable operation of the existing pharmaceutical plants in the country. It is always easy to increase the production in the existing pharmaceutical plants by proper planning of the available resources. Pharmaceutical plants are complex and repairable engineering systems comprising of various units, namely, Weighing Machine, Sifter Machine, Mass Mixer, Granulator, Fluid Bed Dryer, Octagonal Blender, Rotary Compression Machine, Coating Machine and Strip Packing Machine. These units are arranged in a series. One subsystem, namely, the Rotary Compression Machine was supported by standby units with perfect switch over device. The remaining eight subsystems were prone to failure. Mathematical modeling was done by using Markov birth-death process and differential equations were developed on the basis of probabilistic approach using transition diagram. To calculate the variation of reliability with respect to time, the matrix method was described by Goyal (2006) and calculations were done with the help of C-program. The steady-state availability was calculated in order to develop decision matrices which provide the various availability levels for different combinations of failure and repair rates of each subsystem. Kang et al. (2008) discussed the matrix-based system reliability method. Goyal (2006) presented the matrix method for computation of time-dependent availability of a bread manufacturing plant.

In recent times, many researchers have discussed the reliability of different process industries using different techniques. Davis (1952) analyzed some failure data to increase the availability and hence productivity of process industries. Osaki (1972) discussed a two-unit standby redundant system with preventive maintenance. Aggarwal et al. (1975) compared the computational time and absolute error in reliability expression derived by various methods. Kontoleon (1980) discussed the reliability of a special redundant system. Singh et al. (1990) discussed the reliability and availability of the fertilizer and sugar industry. Dayal and Singh (1992) reported the reliability analysis of a system in a fluctuating environment. Subramanian and Ananthraman (1994) reported the reliability of a complex system. Michelson (1998) discussed the use of reliability technology in process industry. Raje et al. (2000) examined the availability of a two-unit standby pumping system. Güines and Deveci (2002) studied the reliability of service systems and their application in student office. Chiange and Yuan (2001) reported on optimum maintenance policy for a markov system under periodic inspection. Habchi (2002) improved the method of reliability assessment for suspended test. Jain (2003) discussed N-Policy for redundant repairable system with additional repairmen.

Science and Technology Journal, Reliability Evaluation, Decision Support System, Pharmaceutical Plant, Mathematical Model, Rotary Compression Machine, Redundant System, Strip Packing Machine, Performance Evaluation, Decision Matrices, Decision-Making Process, Plant Management, Maintenance Resources.