IMPROVING PLANT OPERATIONS WITH PERFORMANCE AND CONDITION MONITORING SYSTEMS

摘要

Business operations in the power industry, as with every other industry, require payback for the resolution of process problems. In order to achieve this payback many plant monitoring systems are used such as Performance and Condition Monitoring Systems. Performance Monitoring systems use first principles calculations for a baseline and test performance case, both of these calculations can then be reconciled to give a cost associated with off design operation. Condition Monitoring systems operate on the Advanced Pattern Recognition Algorithm (APR) and can be used to identify slight variations in the performance of a system largely independent of the quality of the inputs. The process to identify deviations is as follows: a predicted value is calculated for every modeled parameter, a difference between the predicted value and actual value is calculated, the difference is compared to an allowed threshold, and then problems are reported. The purpose of the following report is to identify the strengths and weaknesses of each monitoring system and show how they may be used together for a more thorough analysis of off design operation. Performance monitoring systems provide a reliable and actionable assessment of Heat Rate deviations when they occur in the field. The calculations provided by these systems can lead directly to the diagnosis of real performance problems as well as instrument inaccuracies and provide the financial implications of each issue. Performance monitoring systems, however, are highly dependent upon input quality as their results are found using first principles calculations. Condition Monitoring systems can be used to identify smaller deviations from normal at an earlier time. The major strength of the APR process is its ability to identify these deviations regardless of input quality. As with Performance Monitoring Systems the APR process will identify real problems as well as instrument problems. However, it will not provide a financially quantifiable result as to the effect of the deviation. Monitoring systems should be able to provide quantifiable results withrnminimal personnel use in order to achieve a payback for more operational problems. This paper will discuss how the use of Performance Monitoring Systems in conjunction with Condition Monitoring Systems will provide the complete analysis needed by the power industry today