Condition of cellulose insulation in oil immersed transformers after factory acceptance test

Introduction

Currently, the condition of the solid (cellulose) insulation in a transformer when it leaves the factory is not clearly defined in international standards. It is generally agreed that the insulation shall be new and dry, but there are no numerical limits in the standards, either concerning DP (minimum or average) or concerning remaining moisture (maximum water content). It is common practice to set an initial DP, the DP₀ (the DP after successful factory acceptance test (FAT)) in the range of 800 - 1000. Such values, however, are based on practice rather than on numerical calculations and they are often not accompanied by a moisture content value. In addition, compensation for not meeting DP₀ minimum limits could become a point of dissention between transformer supplier and buyer.

It has to be recognised that the precise determination of the DP and of moisture content of the insulation, even under controlled conditions at the transformer factory, is technically challenging and requires specific expertise. When the transformer leaves the factory, the DP₀, and particularly the moisture content, can vary substantially within the whole insulation system. The wide range of these variations is mainly due to the different types of insulation (from very thin paper to very thick, laminated solid pressboard plates) and to the complex active part construction and conditioning (drying) of power transformers.

Power transformer active parts are processed at intermediate manufacturing stages and there is no uniformity in the final processing during vapor phase drying (VP), or other drying methods, and final oil filling process before testing and hence DP₀ values when measured will show different results.

The present report covers the selection of relevant acceptance parameters, simulations of the relative influence of these parameters and suggested acceptance criteria. It also covers how to measure these parameters. Focus has been to define parameters that are relevant in terms of a direct effect on the transformer insulation lifetime, that are measurable and where achievable limits can be defined and set.

In this study, the effect of the combination of the DP₀ and the initial moisture content of the solid insulation on expected lifetime was investigated in detail, considering specifically the different temperature regimes (conductor paper wrap at the hotspot versus thick laminated pressboard typically at the top and bottom of windings). The brochure provides models for simulating these effects. The determination of the moisture level at delivery is based on the average moisture content of the total insulation. Methods for evaluation of the average moisture are presented. Simulations, as presented in the brochure, are very useful for studying the effect of moisture and DP₀ on the insulation lifetime as well as for defining acceptance limits for these parameters.

Multiple drying processes can sometimes be required, for example following manufacturing or factory acceptance test problems. This issue has been thoroughly investigated by various simulations, including the differentiation of thermally upgraded versus standard Kraft paper and various factory climatic conditions.