Comparative Method for Assessing Cellulose-Based Paper Degradation Rate Immersed in Different Fluids

A transactions article on Power Delivery published by William J. McNutt in 1992 [1] was a milestone for the insulation thermal life discussions. Reviewing historical data starting back in 1916, the author starts his article clarifying the differences between the transformer functional life and the estimation of the insulation life based on different mechanical properties. Clearly stating the scope of the article is the temperature/time relationship to compute life consumption of insulation. After explaining the main degradation mechanisms for cellulose-based materials, where he emphasizes the relevance of the sealing system for having mostly the operating temperature as the main degradation parameter, he recovers the equation proposed by Dakin in 1948 [2], often called the Arrhenius reaction rate equation, with its A and B empirical constants, as a general agreement of the industry. Values of the B parameter of the Arrhenius equation, typically named as the "slope", are presented, ranging from 9,820 to 18,000. After explaining the inherent reasons for such large variation, he proposes the adoption of a single value of 15,000 for all cellulose based materials, regardless of the end-of-life criteria adopted.McNutt’s article presented the fundamental concepts included in the IEEE loading guide C57.91 [3]. During the subsequent three decades, researchers from many different entities around the globe performed additional tests for estimating the insulation life, both using the sealed vessel test procedure and the functional life test (Lockie test). With the introductions of thermally upgraded paper and new solid and liquid insulation materials, new rounds of materials testing were performed. Unfortunately, there is an impression that some of the fundamental concepts seem to be fading away.For the validation of the use of natural ester liquids in the insulation system of distribution and power transformers, some of the historical data had to be recovered, almost developing over again the testing procedures. Amidst this process, different entities and institutions published studies for the development of aging curves when aging both neutral kraft and thermally upgraded paper immersed in natural ester liquids, based on different aging temperatures. Natural ester liquids significantly reduced the degradation rates of the paper in all cases. As described previously by McNutt in the 90’s, the values of A and B parameters varied for each paper testing round.After reviewing some foundations of insulation thermal life, this article will present the alternative comparative methodology developed by the experts who prepared the informative Annex B of IEEE C57.154-2012 [4].Aiming to offer a way to combine several rounds of sealed vessel tests in a single analysis, a procedure for normalizing the results with reference to the "industry proven system" was developed and is presented here. This way, rather than "averaging" the resultant coefficients, the results of aging curves from different entities, performed at different temperature, were combined into a single chart using the horizontal axis as the number of "IEEE units of life". The enriched data fitting procedure leads to a multiplier factor for the degradation rate which can be used for the deployment of new parameters for the Arrhenius reaction rate equation.