Characterisation of Mineral Transformer Oil.

Wilson, Gordon. (2001) Characterisation of Mineral Transformer Oil. Doctoral thesis, University of Surrey (United Kingdom)..

Preview

Text 27750471.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (53MB) | Preview

Abstract

In recent years the electrical industry has begun to address the consequences of the environmental impact and the reduction in physical size of modem high-voltage equipment (such as transformers) on the use of mineral oil. Refiners have recently switched to new sources of crude to produce low-molecular weight, highly refined oils to insulate and cool transformers. To meet these pressures and changes, the National Grid Company (NGC) has begun a programme of analysis to characterise oils from various sources and to attempt to link changes in their composition and physical properties with their deterioration. Detailed analysis allows a more discerning choice among available oils; identification of the changes which occur before breakdown may provide an early warning of such difficulties. The method chosen for fingerprinting transformer oil was gas-chromatographic (GC) analysis of the PAH (poly-aromatic hydrocarbon) fraction. The PAHs were extracted by a liquid-liquid and solid phase extraction (SPE) technique developed after exhaustive experimentation using different SPE cartridges. Unused transformer oils from different sources were used to demonstrate that the PAH fraction was sufficiently characteristic for its use as a source recognition method. Accelerated ageing experiments were carried out to produce degraded oils under similar conditions to those found in large, power transformers. The oils were exposed to high temperature, high voltage and ultraviolet radiation. In each case the PAHs were extracted from the aged oils and analysed by GC. The breakdown of transformer oil is known to produce acidity. Identifying the specific acids produced and monitoring their production in service would be an effective condition assessment tool. Artificially degraded oils with relatively high acidity values, measured by a standard technique, were analysed to identify acidic compounds. A relatively new technique using array-based sensors, commonly referred to as electronic noses, was assessed as a tool for fingerprinting transformer oils.

Item Type:	Thesis (Doctoral)
Divisions :	Theses
Authors :	Wilson, Gordon.
Date :	2001
Additional Information :	Thesis (Ph.D.)--University of Surrey (United Kingdom), 2001.
Depositing User :	EPrints Services
Date Deposited :	14 May 2020 15:43
Last Modified :	14 May 2020 15:49
URI:	http://epubs.surrey.ac.uk/id/eprint/856881

Actions (login required)

View Item

Downloads