This appendix describes the analysis process employed on the available VLF Tan δ Monitored Withstand data to develop criteria for making the condition assessment for filled cable systems.
Also discussed are the test cases used to validate the approach.
Results of the cluster variable analysis of the diagnostic features used to characterize the “Hold”
phase for filled insulations appear in Figure 26. Note that the same feature set as was used in Appendix A is used in Figure 26.
Spd 10-15 (E-3/min) SPD 0-5 (E-3/min)
SPD 0-tfinal (E-3/min) SPD 5-10 (E-3/min)
STD (E-3) Final TD (E-3)
Init TD (E-3) Mean TD (E-3)
7.07
38.05
69.02
100.00
Diagnostic Features
Similarity Level [%]
85.00
Selected Feature
Cluster 1 Cluster 2 Cluster 3
Cluster 4 Cluster 5
Figure 26: Cluster Variable Analysis Results for “Hold” Phase Features for Filled Insulations (Based on data as described in Table 4)
Using the same approach as was used for PE-based insulations, the dendrogram can be reduced to four clusters. In fact, the same features are identified for Filled insulations as PE-based insulations.
Applying PCA to the filled insulations database yields the principal components shown in Table 20.
This table shows the percentage of variance accounted for by each principal component as well as the diagnostic features that contribute the most to each component. Results from Table 20 indicate that only three principal components are required to describe approximately 96% of the data variance.
Table 20: PCA Variances and Component Composition for Filled Insulations
Principal Component
Variance Described by
Component [%]
Variance Described by Component
Cumulative [%]
“Hold” Phase Tan δ Diagnostic
Features
PC1 51.8 51.8 SPD 0-tfinal and STD
(Trend and Variability)
PC2 25.9 77.7 Mean TD
(Level of Loss)
PC3 18.3 96.0 SPD 10-15
(Trend)
PC4 3.7 99.7 Not relevant since these
components only describe 4% of the variability
PC5 0.3 100.0
Figure 27 shows the combined PCA distance of the three principal components for all the available filled insulation Monitored Withstand data. The approach is again the same as the PE-based insulation example in Appendix C. However, the features and feature order used is quite different.
Furthermore, the distances that correspond to the 80th and 95th percentiles are quite different at 0.4 and 2.7, respectively.
10 1
0.1 0.01
99.9 99
90 80 70 60 50 40 30 20
10
5 3 2
1
PCA Distance - Arbitrary Units
Percentage [%]
95 80
0.24 2.7
Figure 27: Empirical Cumulative Distribution of the PCA Distance used for Evaluation of the
“Hold” Phase for Filled Cable Systems
Cable Diagnostic Focused Initiative (CDFI) 10-64
Table 21: Cases Studies for “Hold” Phase Evaluation for Filled Insulations
Case No. Description SPD 10-15 [E-3/min] [E-3/min] SPD 0-5 SPD 0-t[E-3/min] final [E-3] STD
Mean Tan δ [E-3]
Percentage Rank
[%]
1 New System 0.002 0.002 0.002 0.01 3.5 2.5
2 Features at 80% level and
Pos. Speeds * 0.060 0.060 0.060 0.30 13.0 77.0
3 Features at 80% level and
Neg. Speeds * -0.060 -0.060 -0.060 0.30 13.0 79.0
4 Features at 95% level and
Pos. Speeds * 0.600 0.600 0.600 5.00 105.0 94.0
5 Features at 95% level and
Neg. Speeds * -0.600 -0.600 -0.600 5.00 105.0 94.0
6 Utility Test 1 -0.040 -0.040 -0.033 0.30 5.3 72.0
7 Utility Test 2 -0.520 -0.100 -0.287 1.70 22.8 93.0
8 Utility Test 3 1.680 1.120 0.470 5.20 130.7 96.0
* The 80% and 95% diagnostic feature levels correspond to the level of the diagnostic features for
“Hold” phase Evaluation – Decision 2 – Amend Test Time? as shown in Table 9 considering constant speed values during the period under evaluation.
In Table 21, the following examples are included:
Case 1: New Filled cable system lies at the 0.03st percentile. This translates to an extremely good “Pass” margin. Case 1 is represented in Figure 28 by the solid black circle symbol.
Case 2: All diagnostic features set at their respective 80% levels (black square symbol in Figure 28) with positive speeds. Note here that all the features at the 80% level yield a percentage of 77.0%. Therefore, there is a good correlation between the feature levels and the overall assessment considering all features together.
Case 3: All diagnostic features set at their respective 80% levels with negative speeds. Note here that all the features at the 80% level yield a percentage of 79.0%. Therefore, there is again good correlation between the features levels and the overall condition assessment.
Case 4: All diagnostic features set at their 95% levels (black triangle symbol in Figure 28) with positive speeds. In this case, the percentage is 94%. This implies good correlation between the feature levels and the overall assessment considering all features together.
Case 5: All diagnostic features set at their 95% levels with negative speeds. In this case, the percentage is again 94.0%. There is once again good correlation between the feature levels and the overall condition assessment.
Case 6: Real case that represents one of the mid to high performers. The PCA indicates that the cable system is within the mid to higher “No Action” category with a rank of 72.0%.
Case 7: Real case that represents one of the mid to high performer. The PCA indicates the cable system is within the “Further Study” category with a rank of 93.0%.
Case 8: Real case that represents one of the poorest performers (black diamond symbol in Figure 28). The PCA indicates that the cable system is within the poorest 4% of all filled insulated cable systems.
The symbols in Figure 28 represent selected test cases used as examples and their computed PCA distance (Percentage) results appear in Table 21.
10 1
0.1 0.01
99.9 99
90 80 70 60 50 40 30 20
10
5 3 2
1
PCA Distance - Arbitrary Units
Percentage [%]
95 80
New System - Case 1
All Features at 80% level and Pos. Speeds - Case 2 All Features at 95% Level and Pos. Speeds - Case 4
Utility Test 3 - Case 8
Figure 28: Empirical Cumulative Distribution of the PCA Distance used for Evaluation of the
“Hold” Phase for Filled Cable Systems with Relevant Case Studies Presented in Table 21
Cable Diagnostic Focused Initiative (CDFI) 10-66