Underground » Detection and Prevention of Fires and Explosions
This report details the accelerated ageing study carried out during the first stage of Project 4031. Also presented are plans for the next stage of the project. The first element of the project was a survey sent to 60 mines to gain access to the body of experience contained within the coal mining industry. This was followed up by absolute measurements and interviews undertaken by Pacific Power International staff. The results of this study were reported in "Report on Survey of Mine 6.6/11kV Plug and Adaptor Problems 1995".
The report strongly indicated moisture and poor cable joint quality were primary causes in most failures. Pacific Power International's experience with High Voltage Systems suggested that the most likely cause was moisture through mechanisms of partial discharge, (a type of spark erosion).
An environmental chamber was constructed to subject a number of plugs to high levels of moisture, in conditions that would support and promote partial discharges. These conditions were such that in 3 months an accelerated electrical ageing effect, equivalent to approximately 10 years service, was achieved.
This report details the results of the first accelerated ageing test and indicated in general that energised unconnected adaptors were not likely to be involved in failures. The test also indicated precautions that needed to be observed before adequate measurements can be made using conventional techniques.
A great deal of experience has been obtained with the adaptors and the test and measurement system. The environmental chamber is being prepared for the next stage of the test in which an adaptor and plug will be in an in-service configuration, where many crucial design and material elements will be put to the test.
Introduction
This accelerated ageing study was carried out to determine the cause and parameters for the failure of 6.6/11kV underground mining plugs and adaptors. The test objects were subjected to a voltage equivalent to 11kV operation at a frequency of 2kHz for a period of 3 months whilst being maintained in an environment of 90-95% relative humidity.
A description of the study and the results is included in the report, together with details of the next stage.
Results
The partial discharge results pointed to the need for a closer investigation of No. 11. Its partial discharge, DDF (dielectric dissipation factor) and capacitance were measured to determine condition monitoring parameters associated with the discharging.
A standard partial discharge test was carried out on No. 11 and this resulted in the measured value of partial discharge decreasing from 100pC to 50pC and then to 10pC over the course of a 20 minute period. This suggests that the partial discharges were associated with surface moisture which dried during the test. Also after 20 minutes, the DDF measured 32mW/var and the capacitance was determined to be 57pF, both values tending towards their pre-ageing values.
Note, the higher value of PDs logged b the automatic system as compared to the manual test is due to the time it took to set the test up. This allowed conditions on the surface of the adaptor to change so that they were no longer representative of the state within the environmental chamber.
For confirmation, the test was repeated with adaptor 17 which was quickly removed from the environmental chamber and tested. The results were, DDF of 100mW/var and capacitance of 66pF. The plug was left to sit for 40 minutes and the values measured were, DDF 94mW/var and capacitance 65pF and a partial discharge level of <1pC.
To ascertain the distribution of the moisture the plug was dried for 5 minutes with an industrial hot air gun and then retested giving a DDF of 67mW/var and a capacitance of 64pF.
Observations
The results lead to the following observations:
Surface moisture has a strong influence on Partial Discharge, since when the surface moisture is removed the discharges disappear. The ability of the adaptors to return to their pre-test PD values suggests they have not been degraded and are suitable for reuse in the next series of tests.
- Moisture has penetrated into the bulk of the insulation and because the dielectric constant of the water is so high (ยป 60) and water molecules are polar in nature, there is a significant increase in the capacitance and loss angle.
- The reduction of DDF and Capacitance after removal from the humidity controlled room indicates that the moisture penetration is confined to regions near the surface. Otherwise, removing the adaptor from the humidity controlled room for a short period would not show such a significant change in measured values of DDF and capacitance.
- The penetration of moisture into the insulation will simply be a function of time spent in the humidity room. In this case only three months, the accelerated life testing would not have accelerated the moisture ingress rate.
Implications
These test results suggest that if an adaptor has been in a high humidity environment for a long period of time, then the moisture level can be inferred from the DDF measurements. To achieve these measurements in the field would require specific test techniques or an alteration of the plug design. A visual inspection of the adaptors currently show no indication of this deteriorated condition.
Whilst the Partial Discharge effects may eventually cause failure over the surface of the insulation, the general deterioration of the moulded plastic body due to the ingress of water may also contribute the adaptor/plug failures. If water ingress is in fact the predominant failure mechanism, then the environmental conditions will have to be changed to produce accelerated moisture absorption.
Next Stage
A single adaptor, configured in the manner in which it was during this test, can only be considered as a baseline investigation, enabling experience to be gained in both the adaptor material characteristics as well as with the characteristics of the test and measurement system. In its normal configuration (fully assembled) other issues related to clearances and material properties become important. As this assembly will be considerably larger than the previous arrangement there will be a need for a reduction in the number of plug/adaptor assemblies. But in order to maintain a statistically significant sample the hardware will be modified to allow for as many test objects as possible.
The variations in measurement parameters with moisture will require a detailed investigation of the adaptor plastic. To this end, bulk samples of the adaptor plastic will be sought from the manufacturer for a detailed investigation of its hydroscopic and electrical properties.