Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
175
IV. RESULTS that there is no clear cut boundary between electrical faults of
After fault mapping on the thirty five triangles, only three low (D1) and high (D2) energy discharges, and the same applies
emerged as possible fault identification candidates worth to the thermal faults of temperatures T1 & T2 and T3. This is
consideration. These are: practical, since it is not possible to have a clear boundary
Tr1 triangle with the axis; hydrogen, acetylene and ethylene; involving two faults produced by similar stresses (thermal or
electrical). Thus a region of overlap whose size varies from
Tr2 triangle with the axis; methane, acetylene and ethylene
triangle to triangle is real. Triangle Tr3 as visualized is not
which is the Duval Triangle 1;
suitable for identifying partial discharge and thermal faults (T1
Tr3 triangle with the axis; acetylene, ethylene and ethane.
& T2) when compared with the first two. However its good for
The fault mapping on each of the three triangles Tr1, Tr2, Tr3 is
identifying high and low energy electrical discharge faults.
as shown in Figs. 1, 2 and 3 respectively.
From Figs. 1, 2 and 3, it is visually clear faults fall into three Fig. 4. Equipment category and fault type mapping triangle.
distinct regions that are characterized by partial discharge
faults, electrical faults and thermal faults. It can be observed
176
V. DISCUSSION mapping, faults map randomly. Maybe this can be attributed to
From the triangular fault identification technique, the standard the existence of independent mechanisms of their generation
from the oil/paper insulation decomposition. The possibility of
deviation (SD) of each fault category in triangle Tr1, Tr2 and Tr3
alternative Duval triangles 4 & 5 and pentagons will tried in
is as given in Table II. future should field data become available. Finally it is evident
from Table IV that triangular method has higher accuracy
Table II. Standard deviation of faults in Tr1, Tr2, and Tr3. relative to IEEE key gas and IEC 60599 gas ratio methods.
Item PD D1 D2 T1 & T2 T3 Mean
Tr1 0.0371 18.7612 11.5235 21.9404 12.0175 12.8559 ACKNOWLEDGMENT
Tr2 0.3072 14.3541 9.3797 13.2753 9.9989 9.4630 The research was financed by Tshwane University of
Tr3 12.8026 12.4473 7.4727 20.4448 10.2038 12.6742
Technology. Appreciation also to Jomo Kenyatta University of
Technology- Kenya for granting Mr. G. K. Irungu study leave.
Table II shows that PD faults are best located by Tr1 followed
REFERENCES
by Tr2 and finally Tr3. D1 faults are best located by Tr3 followed
[1] IEEE Guide for the Interpretation of Gases Generated in Oil-Immersed
by Tr2 and lastly by Tr1 and the same applies to D2 fault. Transformers. IEEE std. C57.104-2008. New York, USA, 2009.
Thermal faults (T1 & T2 and T3) are well identified by Tr2 then [2] IEC Guide to the Interpretation of Dissolved and Free Gases Analysis-
by Tr3 and lastly by Tr1. Based on average SD Tr2 is the best Mineral Oil-Impregnated Electrical Equipment in Service. IEC60599,
triangle followed by Tr3 and then Tr1. The SD of fault location Geneva, Switzerland, 1999.
as per type of equipment is given by Table III. [3] M. Duval & A. dePablo, Interpretation of gas-in-oil analysis using new
IEC publication 60599 and IEC TC10 data bases, IEEE Electr. Insul.
Mag., Vol. 17 No. 2, pp. 31-41, March/April 2001.
Table III. Standard deviation of fault location as fault and equipment category.
[4] I. A. R. Gray, A guide to transformer oil analysis, Transformer
Equipment I P U R C S B Chemistry Service. [Available]: http://www.satcs.co.za.
Fault [5] G. K. Irungu, A. O. Akumu, & J. L. Munda, Transformer Condition
PD 0.49 - - - - - - Assessment using Dissolved Gas Analysis, Oil Testing and Evidential
D1 - 6.49 10.31 3.61 - 9.43 1.88 Reasoning Approach, Proceedings of 33rd EIC 2015, Seattle , USA, pp.
D2 - 8.69 8.24 7.94 - 7.54 - 145-149, June 2015.
T1 & T2 7.33 10.12 23.10 14.61 - - 26.87 [6] Z. Mao and J. Wen, Detection of Dissolved Gas in Oil-Insulated
T3 - 10.30 10.39 7.98 - - - Electrical Apparatus by Photoacoustic Spectroscopy, IEEE Electr. Insul.
Mag., Vol. 31 No. 4, pp. 7-14, July/August 2015.
Fig. 4 and Table III show that fault location on the triangle does [7] W. H. Tang & Q. H. Wu, Condition Monitoring and Assessment of Power
not depend on equipment type but on fault category only. The Transformers using Computational Intelligence, Springer-Verlag limited.
London, UK, 2011.
accuracy of triangular method was compared with other fault
[8] J. Wada, G. Ueta, S. Okabe & T. Amimoto, Method to Evaluate the
diagnostic methods; IEEE key gas and IEC 60599 gas ratios, Degradation of Transformer Insulating Oil-Establishement of the
the results are summarized in Table IV. Here faults are Evaluation Method and Application to Field Transformer Oil, IEEE
classified into four categories as per IEEE key gas method: Trans. on Dielectrics and Electr. Insul., Vol. 22, No.2, pp. 1266-1274.
April 2015.
partial discharge, arcing, thermal faults in paper (T1 & T2) and
[9] N. A. Bakar, A. Abu-Siada and S. Islam. A Review of Dissolved Gas
thermal fault in oil (T3) [1]. Analysis Measurement and Interpretation Techniques, IEEE Electr.
Insul. Mag., Vol. 30. No. 3, pp. 39-49. May/June 2014.
Table III. Fault diagnostic accuracies of different methods.
[10] Transformer: Basics, Maintenance and Diagnostics. US Department of the
Fault PD Arcing T1 & T2 T3 Mean Interior Bureau of Reclamation. April, 2005.
Method accuracy [11] A. Abu-Siada & S. Islam, A New Approach to Identify Power
IEEE Key gas 88.89% 37.84% 50.00% 22.22% 49.74% Transformer Criticality and Asset Management Decision Based on
IEC gas ratios 44.44% 86.49% 62.50% 66.67% 65.03% Dissolved Gas-in-Oil Analysis, IEEE Trans. on Dielectrics and Electr.
Tr1 100% 98.65% 75% 88.89% 90.64% Insul., Vol. 19, No.3, pp. 1007-1012. 2012.
Tr2 100% 100% 87.50% 88.89% 94.10% [12] Transformer Maintenance: Facilities Instructions, Standard, and
Tr3 88.89% 98.65% 87.5% 72.22% 86.82% Techniques, FIST 3-30. US Department of the Interior Bureau of
Reclamation. October, 2000.
[13] M. Duval & L. Lamarre, The Duval Pentagon-A New Complimentary
VI. CONCLUSIONS Tool for the Interpretation of Dissolved Gas Analysis in Transformers,
IEEE Electr. Insul. Mag., Vol. 30 No. 6, pp. 9-12, Nov/Dec. 2014.
A review of fault identification in oil filled electrical [14] M. Duval, Use of Pentagons and Triangles for the Interpretation of DGA
equipment using Duval triangle 1 and other possible triangles in Electrical Equipment, Proc. Of TechCon North America Conf.,
has been given. Out of the possible thirty five triangles only Albuquerque, Feb. 2016.
three proved to be reasonably useful in fault identification. It has [15] M. Duval, A Review of Faults Detectable by Gas-in-Oil Analysis in
been shown that based on average standard deviation Duval Transformers, IEEE Electr. Insul. Mag., Vol. 18 No. 3, pp. 8-17,
triangle 1 (Tr2) is the best in fault identification among the three May/June 2002.
triangles although not in all fault categories. It can be concluded [16] M. Duval, State of the Art of Technical Diagnosis through Intepretation
of DGA, Proc. Of My Transfo Conf., Torino, Italy, Nov. 2012.
that equipment type does not have any effect on the fault location
as visually noted and confirmed by standard deviation statistics. [17] D. Graham & N. Midgley, Tri-plot software documentation. 2003.
[Available]: http://www.Iboro.ac.uk/research/phys-geog/tri-plot.
In addition, it was seen that triangles with axies of hydrogen or
hydrocarbon and carbon oxide gases cannot be used for any fault
177