ABSTRACT
It has long been known that comparing the partial discharge results obtained from a single machine is a valuable tool enabling companies to observe the gradual deterioration of a machine stator winding and thus plan appropriate maintenance for the machine [1]. In 1998, at the annual Iris Rotating Machines Conference (IRMC), a paper was presented that compared thousands of partial discharge (PD) test results to establish the criteria for comparing results from different machines and the expected PD levels [2]. At subsequent annual Iris conferences, using similar analytical procedures, papers were presented that supported and expanded upon the previous criteria [6 – 21]. This paper presents the most recent statistical analysis of the database.
In addition to presenting the most recent analysis of the database, in this paper we also reviewed the influence of winding impregnation method on the PD levels. There are three primary manufacturing processes used for resin penetration: resin-rich (B-stage) tapes, batch vacuum pressure impregnation (VPI) and global VPI. Over the past several years, the global VPI process has been more widely used as it accelerates the manufacturing process. Since the presence and magnitude of PD are well-established indicators of resin penetration and insulation consolidation, this paper compares the PD results for non-global and global VPI processes. Previous papers have determined that voltage and test configurations highly influence the results, thus the comparisons are made for assets within the same voltage classes and data gathered using identical test specifications.