Medium Voltage (MV) extruded power cables play a vital role in electric power transmission and distribution systems. Insulation of power cables is essential for cables’ lifetime, effectiveness, and reliability within the power cable system. However, insulation materials degrade progressively due to thermal, electrical, mechanical, and environmental stresses experienced during cables operation. This ageing of power cables poses potential safety concerns for power system operators and leads to increased maintenance and replacement costs.
This research focuses on the thermal ageing of MV extruded power cables. In order to investigate electrical performance degradation of entire extruded power cables (comprising semi-conductor layers) under thermal ageing, there are four single-core extruded MV cables with cross-linked polyethylene (XLPE) insulation are subjected to accelerated thermal ageing in a chamber. Standard measurements of Insulation Resistance (IR), Polarization Index (PI), Dielectric Absorption Ratio (DAR) and Dielectric Loss (tanδ) are carried out on the specimens periodically through the thermal ageing process. It was found that IR and tanδ measurements are susceptible to Relative Humidity (RH), while PI variations are weak relative to RH changes. In addition, the annealing effect, which existing in extruded cable with semi-conductor layer through thermal ageing processes, has a significant influence on IR and tanδ.
To better evaluate degradation of cable insulation caused by distributed thermal stresses, a methodology combining Finite Volume Method (FVM) and Artificial Neural Network (ANN) models are developed to determine spatial temperature profiles of the cable insulations in the experimental environment. The thermal modelling temperature distribution results from FVM provide temperature distribution on cross-section of cables and align well with the experimentally measurement results. A much more detailed and reliable temperature distribution profile of cable insulation estimated from ANN model, which is trained by FVM resulting data, holds significant benefits for estimating thermal degradation of extruded power cables.
IR degradation models for extruded power cables under thermal ageing are established by using of dichotomy and discretization models. Cable cylindrical insulation is first divided into sufficiently small segments, and then simulated by dichotomy models that randomly sample fully degraded segments based on an overall (or layer) ageing condition estimation and discretization models that estimated the gradual degradation of individual segments, respectively. Furthermore, uniform and non-uniform temperature profiles are incorporated into dichotomy and discretization models, respectively, for a comparison. The IR simulation results are not only compared between different models, but also discussed around the sensitivity of IR simulation to segment size and degradation rates. Among the four developed IR degradation models, the discretization model with non-uniform temperature distribution is recommended, as it more comprehensively accounts for the effects of temperature distribution during the aging process and treats the IR of the material as a bulk property. However, the limitation of this model is its need for more accurate degradation rates as a function of aging temperature.
Finally, in accordance with IEEE Standard 1407, Finite Element Method (FEM) thermal modelling works of power cable accelerated ageing in a tank filled with water is proposed. Simulation investigates there are temperature gradients across cable length and temperature difference among cables in tank, which are influential for ageing rate of cables in experiment based on IEEE standard 1407. To achieve a uniform temperature distribution, a forced circulated system with different rotating speed and at different location in a tank are simulated and compared. Based on the simulation results, it is highly recommended that IEEE Standard 1407 include guidelines for using forced water circulation with a fan speed of 50 rpm to create a more uniform artificial aging environment for all cables in the tank.
Date of Award | 24 Sept 2024 |
---|
Original language | English |
---|
Awarding Institution | - University Of Strathclyde
|
---|
Sponsors | University of Strathclyde |
---|
Supervisor | Brian Stewart (Supervisor) & Martin Given (Supervisor) |
---|