Detailed procedures of retrofilling transformers with FR3 natural ester and residual mineral oil content testing

Abstract

Several retrofillings have been performed by different utilities in China. However, there is still insufficient understanding of the retrofilling procedures for minimising residual mineral oil and on the effect of the temperature rise of transformers. Following new detailed procedures, an out‐of‐service S11‐M‐315/10 (10 kV 315 kVA) and a new S13‐M‐400/10 (10 kV 400 kVA) distribution transformers were selected to retrofill their mineral oil with FR3, a soybean‐based natural ester. A method was proposed to predict the residual mineral oil content, which uses the kinematic viscosity at 60°C. This method fills a gap in the testing of residual mineral oil content. The mass ratios of residual mineral oil were 1.24% for the S11‐M‐315/10 transformer and 2.05% for the S13‐M‐400/10 transformer, based on the procedures in this study. The fire point of the mixed oil after retrofilling was 352°C, which meets the K‐class less‐flammable requirements, indicating a proper control of residual mineral oil. After retrofilling, the time of reaching the limits of the 1.5 times rated load was 30% longer than that of the mineral oil transformer. Two retrofilled transformers have been operating in good conditions since November 2019 for the S11‐M‐315/10 transformer and December 2020 for the S13‐M‐400/10 transformer.

A New Method for the Quantification of the Remaining Mineral Oil in Natural-Ester Retrofilled Transformers

“…McShane [13] and Wang [14] 4-7% Fofana [10] and Toudja [12] 10% Beroual [11] 10-20% As stated above, this work provides an accurate method to quantify the residual MO that remains in a transformer after retrofilling it with NE. The method can be used to assess the quality of the process and to support the interpretation of predictive maintenance measurements on the transformer after the retrofilling.…”

“…Some studies provide estimates on the amount of residual MO that mixes with the ester after a retrofilling process which differ significantly between them (Table I). Most authors give no information about the methods used to obtain the values, making it difficult to assess their reliability; only Y. Wang et al [14] specify that their estimation of the percentage of contamination rate is based on viscosity measurements.…”

“…Quantifying the amount of residual MO that mixed the NE after a retrofilling is important to assess the quality of the process and to make sure that the fire point is not reduced to below 300ºC. It is known that a MO content higher than 7% dramatically decreases the fire and flash point of NEs [13], [14]. Estimating the MO content is also important for maintenance and to get insight into the dielectric performance of the oil-paper insulation system.…”

“…Estimating the MO content is also important for maintenance and to get insight into the dielectric performance of the oil-paper insulation system. Some authors have proposed methodologies to quantify the contamination of a transformer with MO based on the measurement of several properties of the oil such as relative permittivity [15] or kinematic viscosity [14]. However, these methods are not precise or reliable enough to obtain an accurate estimate of the percentage of MO dissolved in the NE (i.e oil permittivity is influenced by other factors such as its water content or degree of oxidation, and the range of variation of density or viscosity is not wide enough to get precise estimates).…”

Life Expectancy of Transformer Paper Insulation Retrofilled with Natural Ester in the Laboratory

“…Although there is agreement on the fact that cellulose insulation ages more slowly in the presence of an ester, it is not so clear that this result can be extrapolated to the insulation of retrofilled transformers. Experimental evidence on this fact is mainly based on laboratory accelerated ageing tests carried out on cellulose-ester systems, but the liquid insulation of retrofilled transformers is a mixture of mineral oil (MO) and natural (NE) or synthetic ester [7][8][9], the MO being the material that remains mostly adsorbed in the paper [10,11].…”