Complex systems are subject to failures because of many reasons which affect the profitability of the manufacturing industry and hence reliability analysis plays an important role in understanding the system performance while dealing with real industrial problems under different operating conditions and assumptions. This PhD research focuses on conducting reliability analysis of rodding anode plant in aluminum industry. Six years maintenance data on component failures, repairs and associated costs are collected for the purpose of this analysis and various rates are estimated from the data. The Semi-Markov processes and regenerative point techniques are used in the analysis in an attempt to measure system effectiveness such as mean time to plant failure (MTPF), availability of the plant, busy period of repairman and the number of repairs. The expected outcome of this research study is to estimate reliability to enable the plant engineers understand the plant failure pattern and thereby built up a scope for improvement in terms of plant performance by reviewing maintenance strategy.