International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 1
ISSN 2229-5518
The life of a transformer is dependant upon life of its insulation. The insulation of a transformer deteriorates over a span of time with temperature, moisture and oxygen. Oil quality assessment is the technique to decide the quality of transformer oil after a specific period of use. The quality of transformer oil decides the efficiency of the transformer while in service. This chapter presents an innovative methodology for assessment of transformer oil quality by using fuzzy logic based decision support system.
The quality of transformer oil can be termed as good if it fulfills the following functional requirements for at least 30 years of service.
It acts as an electrical insulator.
It acts as a coolant for transformer winding and core.
Transformer oil free from water particles and sludge acts as a good insulator. Gases are produced in a transformer when it is subjected to electrical, thermal and environmental stresses. These gases get dissolved in oil and provide information about the incipient faults developing in transformer. The oil quality assessment is necessary for deciding preventive maintenance schedules of transformers. So far during the course of study of literature, no work has been found on the assessment of oil quality for deciding future course of action, after a specific period of use. In this dissertation work, an attempt is being made to assess oil quality by using fuzzy logic to arrive at a
decision regarding the action to be taken.
IJSER © 2012 http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 2
ISSN 2229-5518
There is a tendency of transformer oil to absorb water vapour from atmosphere due to breathing during service and even during transportation and storage. Water is injurious to transformer insulation system due to the following reasons [1].
It reduces the electric strength of oil.
It reduces the resistivity of oil.
It accelerates deterioration of solid insulation.
Air dissolved in oil induces risk of bubble formation and accelerated oxidation process, leading to chemical deterioration of oil. The biggest contaminator, moisture enters transformer oil from following sources.
by leakage past gasket
by absorption within the transformer as a product of degradation of insulation at high temperatures
Transformer oil also gets impregnated with various types of impurities. These may be solid ones like hygroscopic fibers, suspended particles or liquid ones like dissolved water organic and inorganic gases. They bring about a considerable reduction in the dielectric strength of the oil. As an example, oil at 20oC saturated with water (44 PPM) attains only about 25% of the original electric strength with water content of 10 PPM.
Thus, transformer oil is reconditioned to eliminate the following elements.
free and dissolved water
solid impurities
dissolved gases
The possible reconditioning processes are:
Single Filtering and Degassing
Double Filtering and Degassing
Reclamation
IJSER © 2012 http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 3
ISSN 2229-5518
The oil quality assessment method developed in this dissertation is based on finding the Total Hydrocarbon Gas (TCG) content in the oil. Depending upon TCG and severity of different gases present, the Oil Degradation Index (ODI) is evaluated and used to assign one of the above possible courses of action, if it is suitable for reconditioning. The single stage and double stage filtering and degassing are done by vacuum plants. This procedure involves course and fine filtration followed by degassing and dehumidification. The transformer oil is spread over a large surface area under vacuum, after which oil is delivered for use. The single stage and double stage filtering decision depends upon degradation level of used oil as manifested by the concentration of dissolved gases. In addition there are two other classifications.
No Filtering
Do Not Use
“No Filtering” condition means oil is not degraded and can be reused without any requirement of reconditioning. “Do Not Use” condition means whole transformer oil is to be replaced with new one, as the oil is degraded to an extent that its reconditioning has become impossible.
Oil reclamation is a process to remove heterogeneous atoms and oxidized products from oil. This is a chemical treatment of oil followed by filtering and degassing. The aim is to remove acid and other harmful compounds by precipitation, sludge formation or by use of chain reaction of organic compounds.
It is easy to visualize that considerable uncertainty is involved in the process of defining transformer oil quality. Hence, it is imperative to use fuzzy
logic for this purpose. The fuzzy measure for the assessment of oil quality is
IJSER © 2012 http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 4
ISSN 2229-5518
based on a combination of knowledge about TCG along with particular key gases evolved due to fault condition. The decision making process is based on the degree of match between current and permissible limits of dissolved gases and the fuzzy rule based system. The method is designed to compute the fuzzy measure named as transformer Oil Degradation Index (ODI), by integration of the information of different hydrocarbon gases formed in transformer oil. This approach uses a knowledge base to derive Oil Degradation Index (ODI). The transformer oil quality data is transformed into a normalized fuzzy number with membership grade function adjusted for characterizing transformer oil quality based on the data available from both normal and faulty transformers.
The input variables to the present classification have different ranges in various hydrocarbon gases viz. CH4, H2, C2H6, C2H4, and C2H2 developed during incipient fault detection on the basis of DGA and TCG (Total Gas Content). Trapezoidal membership functions (trapmf) are used for all input variables. The term set of membership functions for all input variables except
TCG are:
Low
Moderate
High
Severe
The input variable TCG uses another term set ‘V. High’ in addition to above term sets. The input membership functions along with different term sets are shown in Figure 4.1. This figure is generalized for different inputs. The ranges of membership functions for input variables are shown in Table
4.1.
There is only one output variable to assign transformer oil quality. This
is called Oil Degradation Index (ODI). ODI value is computed for each input
IJSER © 2012 http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 5
ISSN 2229-5518
variable defined above. This ODI value and severity of fault as defined by rule base, decide the process to be adopted for reconditioning. The input membership functions are designed to give a constant output over substantial range. The membership of each input variable varies over a definite range, uniformly on both sides (trapezoidal). The full membership (value 1) covers
80% area of total range. This is the reason why output gives a constant ODI in maximum cases thus, making this measure insensitive to small changes in gas concentrations. This is an extremely useful property to take care of normal changes due to operations and ageing. Another significant feature is that this ODI is further correlated to the specific reconditioning procedure of transformer oil. The output membership functions are shown in Figure 4.2 and corresponding range and type of membership functions are shown in Table
4.2.
input i
Low
1
Moderate
High V. High Severe
input i
IJSER © 2012 http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 6
ISSN 2229-5518
ODI
MF 1
1
MF 2 MF 3 MF 4 MF 5
1
ODI
Note:
MF 1 = No Filtering
MF 2 = Single Filtering & Degassing MF 3 = Double Filtering & Degassing MF 4 = Reclamation
MF 5 = Do Not Use
IJSER © 2012 http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 7
ISSN 2229-5518
The fuzzy rules are developed to match a specific output condition of oil depending upon various input values. 38 rules are developed in all.
Table 4.3 shows the input data for oil quality analysis. The corresponding outputs are shown in Table 4.4. The output oil quality or Oil Degradation Index (ODI) is shown for individual group of inputs.
INPUT DATA | ||||||
S. No. | CH4 | H2 | C2H6 | C2H4 | C2H2 | TCG |
1. | 4 | 3 | 4 | 3 | 0 | 14 |
2. | 2 | 1 | 1 | 0 | 0 | 4 |
3. | 2 | 2 | 0 | 0 | 0 | 4 |
4. | 3 | 2 | 3 | 1 | 0 | 9 |
5. | 2 | 2 | 4 | 3 | 0 | 11 |
6. | 4 | 2 | 4 | 3 | 0 | 13 |
7. | 3 | 2 | 3 | 0 | 0 | 8 |
8. | 20 | 9 | 5 | 10 | 0 | 44 |
9. | 4 | 7 | 3 | 1 | 0 | 15 |
10. | 10 | 14 | 8 | 5 | 0 | 37 |
11. | 15 | 13 | 4 | 3 | 0 | 35 |
12. | 10 | 13 | 2 | 1 | 0 | 26 |
13. | 7 | 17 | 13 | 3 | 0 | 40 |
14. | 9 | 12 | 3 | 7 | 0 | 31 |
15. | 1 | 0 | 0 | 0 | 0 | 1 |
16. | 125 | 95 | 16 | 281 | 0 | 517 |
17. | 212 | 68 | 54 | 470 | 0 | 804 |
18. | 107 | 32 | 17 | 265 | 0 | 421 |
19. | 31 | 42 | 16 | 115 | 0 | 204 |
20. | 144 | 60 | 67 | 449 | 9 | 729 |
21. | 95 | 1076 | 4 | 71 | 231 | 1477 |
22. | 20 | 240 | 5 | 28 | 96 | 389 |
23. | 32 | 338 | 1 | 32 | 50 | 453 |
24. | 117 | 531 | 27 | 132 | 848 | 1655 |
IJSER © 2012 http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 8
ISSN 2229-5518
25. | 71 | 79 | 3 | 72 | 115 | 340 |
26. | 38 | 41 | 3 | 45 | 58 | 185 |
27. | 104 | 110 | 30 | 86 | 131 | 461 |
28. | 248 | 125 | 10 | 147 | 179 | 709 |
29. | 262 | 222 | 27 | 168 | 410 | 1089 |
30. | 91 | 98 | 18 | 72 | 154 | 433 |
31. | 43 | 48 | 3 | 75 | 81 | 250 |
32. | 1107 | 201 | 110 | 2016 | 6350 | 9784 |
33. | 47 | 442 | 117 | 67 | 69 | 742 |
34. | 1417 | 114 | 296 | 2096 | 0 | 3923 |
35. | 4 | 2 | 3 | 4 | 0 | 13 |
36. | 99 | 747 | 13 | 97 | 1036 | 1992 |
37. | 122 | 41 | 31 | 143 | 188 | 525 |
38. | 339 | 59 | 42 | 392 | 1 | 833 |
39. | 4 | 7 | 3 | 2 | 0 | 16 |
40. | 34 | 21 | 5 | 47 | 62 | 169 |
41. | 21 | 199 | 0 | 40 | 144 | 404 |
42. | 5365 | 2973 | 427 | 5532 | 2124 | 16421 |
43. | 61 | 65 | 16 | 143 | 3 | 288 |
44. | 87 | 16 | 75 | 395 | 30 | 603 |
45. | 186 | 813 | 15 | 249 | 1001 | 2264 |
46. | 38 | 212 | 15 | 47 | 78 | 390 |
47. | 16615 | 2754 | 3657 | 31476 | 613 | 55155 |
48. | 1393 | 800 | 304 | 2817 | 3000 | 8314 |
49. | 770 | 199 | 217 | 1508 | 72 | 2766 |
50. | 8784 | 4906 | 1404 | 9924 | 9671 | 34689 |
51. | 13 | 24 | 5 | 43 | 319 | 404 |
52. | 584 | 266 | 328 | 862 | 1 | 2041 |
53. | 10 | 160 | 3 | 1 | 1 | 175 |
54. | 619 | 80 | 326 | 2480 | 0 | 3505 |
55. | 3997 | 231 | 1726 | 5584 | 0 | 11538 |
56. | 24 | 127 | 0 | 32 | 81 | 264 |
57. | 4066 | 9474 | 353 | 6552 | 12997 | 33442 |
58. | 1053 | 507 | 297 | 1440 | 17 | 3314 |
59. | 695 | 416 | 74 | 867 | 0 | 2050 |
60. | 207 | 441 | 43 | 224 | 261 | 1176 |
61. | 38 | 212 | 47 | 15 | 78 | 390 |
62. | 1393 | 800 | 304 | 2817 | 3000 | 7314 |
63. | 770 | 199 | 217 | 1508 | 72 | 2766 |
64. | 8784 | 4906 | 1404 | 9924 | 9671 | 34689 |
65. | 1742 | 425 | 7299 | 37043 | 158 | 62349 |
66. | 95 | 1076 | 4 | 71 | 231 | 1477 |
IJSER © 2012 http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 9
ISSN 2229-5518
67. | 754 | 244 | 172 | 1281 | 27 | 2478 |
68. | 167 | 117 | 48 | 481 | 7 | 820 |
69. | 1324 | 858 | 208 | 2793 | 7672 | 12855 |
70. | 369 | 137 | 144 | 1242 | 16 | 1908 |
71. | 27 | 274 | 5 | 33 | 97 | 436 |
72. | 370 | 1249 | 56 | 606 | 1371 | 3652 |
73. | 20 | 240 | 5 | 28 | 96 | 389 |
74. | 79 | 33 | 30 | 215 | 5 | 362 |
75. | 22 | 307 | 2 | 33 | 109 | 473 |
76. | 144 | 60 | 67 | 449 | 9 | 729 |
77. | 9739 | 2004 | 2750 | 5113 | 0 | 19606 |
78. | 107 | 127 | 11 | 157 | 224 | 623 |
S. No. | Oil Degradation Index (ODI) | Reconditioning Method |
1. | 0.0784 | No Filtering |
2. | 0.0784 | No Filtering |
3. | 0.0784 | No Filtering |
4. | 0.0784 | No Filtering |
5. | 0.0784 | No Filtering |
6. | 0.0784 | No Filtering |
7. | 0.0784 | No Filtering |
8. | 0.0784 | No Filtering |
9. | 0.0784 | No Filtering |
10. | 0.0784 | No Filtering |
11. | 0.0784 | No Filtering |
12. | 0.0784 | No Filtering |
13. | 0.0784 | No Filtering |
14. | 0.0784 | No Filtering |
15. | 0.0784 | No Filtering |
16. | 0.2500 | Single Filtering and Degassing |
17. | 0.2500 | Single Filtering and Degassing |
18. | 0.2500 | Single Filtering and Degassing |
19. | 0.2500 | Single Filtering and Degassing |
20. | 0.2500 | Single Filtering and Degassing |
21. | 0.4500 | Double Filtering and Degassing |
22. | 0.2500 | Single Filtering and Degassing |
23. | 0.2500 | Single Filtering and Degassing |
24. | 0.4500 | Double Filtering and Degassing |
IJSER © 2012 http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 10
ISSN 2229-5518
25. | 0.2500 | Single Filtering and Degassing |
26. | 0.2500 | Single Filtering and Degassing |
27. | 0.3500 | Double Filtering and Degassing |
28. | 0.2500 | Single Filtering and Degassing |
29. | 0.3500 | Double Filtering and Degassing |
30. | 0.2500 | Single Filtering and Degassing |
31. | 0.2500 | Single Filtering and Degassing |
32. | 0.8610 | Do Not Use |
33. | 0.4500 | Double Filtering and Degassing |
34. | 0.6500 | Reclamation |
35. | 0.0784 | No Filtering |
36. | 0.4500 | Double Filtering and Degassing |
37. | 0.4500 | Double Filtering and Degassing |
38. | 0.2500 | Single Filtering and Degassing |
39. | 0.0784 | No Filtering |
40. | 0.2500 | Single Filtering and Degassing |
41. | 0.2500 | Single Filtering and Degassing |
42. | 0.7770 | Do Not Use |
43. | 0.2500 | Single Filtering and Degassing |
44. | 0.4500 | Double Filtering and Degassing |
45. | 0.6500 | Reclamation |
46. | 0.2500 | Single Filtering and Degassing |
47. | 0.7770 | Do Not Use |
48. | 0.7770 | Do Not Use |
49. | 0.6500 | Reclamation |
50. | 0.7770 | Do Not Use |
51. | 0.4500 | Double Filtering and Degassing |
52. | 0.4500 | Double Filtering and Degassing |
53. | 0.2500 | Single Filtering and Degassing |
54. | 0.6500 | Reclamation |
55. | 0.7770 | Do Not Use |
56. | 0.2500 | Single Filtering and Degassing |
57. | 0.7770 | Do Not Use |
58. | 0.6500 | Reclamation |
59. | 0.4500 | Double Filtering and Degassing |
60. | 0.3500 | Double Filtering and Degassing |
61. | 0.2500 | Single Filtering and Degassing |
62. | 0.7770 | Do Not Use |
63. | 0.4500 | Double Filtering and Degassing |
64. | 0.7770 | Do Not Use |
65. | 0.8710 | Do Not Use |
66. | 0.4500 | Double Filtering and Degassing |
67. | 0.4500 | Double Filtering and Degassing |
68. | 0.2500 | Single Filtering and Degassing |
69. | 0.7770 | Do Not Use |
70. | 0.4500 | Double Filtering and Degassing |
71. | 0.2500 | Single Filtering and Degassing |
72. | 0.6500 | Reclamation |
IJSER © 2012 http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 11
ISSN 2229-5518
73. | 0.2500 | Single Filtering and Degassing |
74. | 0.2500 | Single Filtering and Degassing |
75. | 0.2500 | Single Filtering and Degassing |
76. | 0.2500 | Single Filtering and Degassing |
77. | 0.7770 | Do Not Use |
78. | 0.2500 | Single Filtering and Degassing |
The method applied in this work is original and innovative in nature. During periodical incipient fault diagnosis of transformers, Oil Degradation Index (ODI) gives additional information for representing quality of oil. This helps in better monitoring of transformer condition. The output results of oil quality assessment can be used for making appropriate decisions regarding the procedure to be adopted for the reconditioning of transformer oil. This will improve life of transformer oil and life of transformer as a whole. Hence, preventive maintenance schedules can be programmed optimally with minimum breakdown.
REFERENCES
1. Bharat Heavy Electricals Limited, Bhopal, Transformers, Tata McGraw- Hill Publishing Company Limited, New Delhi, 1997.
2. A.C. Franklin and D.P. Franklin, “The J&P Transformer Book”, 11th
Edition, Aditya Book Pvt. Ltd, New Delhi 1995.
3. Joseph B. Digiorgio, “Dissolved Gas Analysis of Miniral Oil Insulating
Fluids”, http.//www.nttworldwide.com/tech2102.htm.
4. J.A. Momoh, X.W. Ma, K. Tomsovic, “Overview and Literature Survey of Fuzzy Set Theory in Power System. IEEE Transactions on Power Systems, Vol. 10, N0-3, August 1995.
5. K.Tomsovic, A.Amar, On Refining Equipment Condition Monitoring
Using Fuzzy Sets and Artificial Neural Nets, International Journal of
Engineering Intelligent Systems, Vol.5, No.1, March 1997, 43-50.
IJSER © 2012 http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 12
ISSN 2229-5518
6. G.N.S. Kalyani D.V.S.S. Sivasarma, “AI Techniques for Condition Monitoring of Power Trasformers Using DGA”. National Power Systems Conference, NPSC 2004, 1110 – 1115.
7. A.G. Levit, O.N. Grechko, N.P. Shchipunova, Allowance for Insulation Ageing in the New Concept of Accelerated Life Tests of High Voltage Power Transformers, IEEE Transactions on Power Delivery, Vol.7, No.3, July 1992.
8. R.M. Radwan, R.M. El-Dewieny, I.A. Metwally, Investigation of Static Electrification Phenomenon Due to Transformer Oil Flow in Electric Power Apparatus,IEEE Transaction on Electrical Insulation, Vol. 27, No. 2, April 1992.
9. IEEE Substations Committee Working Group K4, Partial Discharge Testing of Gas Insulated Substations, IEEE Transactions on Power Delivery, Vol.7, No.2, April 1992.
10.J Aubin, Y. Langhame, Effect of Oil Viscosity on Transformer Loading Capability at Low Temperatures, IEEE Transactions on Power Delivery, Vol.7, No.2, April 1992.
11.W.J. Mcnutt, T.O. Rouse, G.H. Kaufmann, Mathematical Modelling of
Bubble Evolution in Transformers, IEEE Transaction on Power
Apparatus and Systems, vol. 104, No. 2, February 1985.
IJSER © 2012 http://www.ijser.org