Transformer insulation system-a review of the guiding principles for thermal classification

Abstract

material that has been chosen as desirable for other reasons. For example, to what extent must the prevalence of cavities of certain depths in extruded, solidified or liquid impregnated systems be controlled from an economic point of view?The matter of permissible corona intensity raises the question of how sensitive the detection equipment must be to meet the requirements of the test. Historically, the use of corona detection equipment has spread from the research laboratory to quality control on the production floor, and the demand for higher sensitivities has changed from several hundred to less than 5 pico-coulombs. Based on the present "state of the art" this trend is taxing the capabilities of corona-detection-equipment suppliers. Greater sensitivity eventually reduces the ratio of signal to threshold noise. Furthermore, the use of corona detection equipment on the production floor increases the demand for semi-automatic features, such as logic systems, that reduce the need for personnel decisions. All of this modern sophistication increases test-equipment cost. ConclusionsFrom design , production, marketing and end-use points of view it is important that the corona problem, as it relates to electrical devices, be fairly well understood. The object here has been to focus on the four most significant phases of the problem. The principles of the ionization process and stress control are well documented in the literature. Detection-sensitivity requirements, and the corona resistance of insulating materials are related questions that are not as well documented.If the use of more modern insulating materials is desirable for economic reasons, how vulnerable are they to corona attack, IntroductionSince the beginning of this century, by first working with insulation materials, then evaluation of entire insulation systems, a solid foundation of the time-temperature relationship of the deterioration of insulation has been developed. The time-temperature relationship is established by accepted functional accelerated life tests on the actual units or on models. Correlation of the test procedure with field experience for the purpose of predicting thermal classification and acceptable life has been accomplished.There appears to be some differences of opinion as to the interpretation of the data from the functional accelerated life tests. The accelerated life test was designed to be severe to obtain data in a short period of time; and since these data has been correlated with systems that have much field experience under normal field conditions, it would not seem logical to draw conclusions from these data for systems operating under radically different field conditions. For this set of field conditions, it would appear that a different functional accelerated life test should be used from which to draw conclusions regarding the thermal rating and life.