Internal winding faults resulting from the degradation of transformer winding insulation can be catastrophic and hence expensive. In the new environment of deregulation, utilities therefore need inexpensive methods employed to detect such faults in the incipient stage. However, the implementations of the existing monitoring methods tend to cost too much to be applied to distribution transformers. Therefore, an ongoing project in the power systems automation laboratory (PSAL) of Texas A&M University is to develop on-line incipient fault detection methods for single-phase distribution transformers that utilize the terminals parameters of voltages and currents. The development of an accurate internal fault diagnostic technique for transformers must be based on the analysis of quantities from fault scenarios. Considering the safety of personnel, the damage that will occur in the transformer, the consumed time, and related cost, simulation involving the modelling of transformers at various incipient fault stages is the best way to generate these fault cases. This paper presents new methodology developed to model internal incipient winding faults in distribution.Transformers based on the author’s earlier work involving the development of a two-dimensional (2-D) nonlinear finite element analysis internal short circuit model. Since deteriorating insulation involves two stages aging and arcing model and an arcing, the degrading insulation model is composed of an aging model and an arcing model was combined with the internal short circuit model, developed in our earlier work, to simulate internal winding incipient faults. The transformer internal incipient fault model was implemented using commercially available finite element analysis software. Various incipient fault scenarios at different degrading levels of the transformer were analyzed in both time domain and frequency domain. The characteristics obtained from the simulation were compared with the characteristics obtained from some experimental fault cases that conveyed incipient like behaviour.
In this paper, the transformer model to simulate internal short circuit winding faults s briefly introduced in section second. Then the method to simulate an incipient internal winding faults s discussed in more details in section third. In section forth, some simulation results are discussed and compared with experimental results. The conclusions are given in last section.
2. INTERNAL SHORT CIRCUIT FAULT MODEL
A method was developed to apply finite element analysis to calculate the parameters for an equivalent circuit of the transformer with an internal short circuit fault using ANSOFT’s Maxwell software. To model an internal fault on the primary or secondary winding, the faulted winding is divided into two subsoil “a” and “b” (turn-to-earth fault) or three subsoil ”a,” ”b,” and ”c” (turn-to-earth fault).
The primary winding and the secondary windings are represented by polygons of corresponding materials. For instance, to simulate a turn-to-earth fault on the primary winding. The transformer is modelled as shown in the fig. 1(a). The primary winding s divided into subsoil a and b. The transformer model to simulate a turn-to-turn fault between two turns on the primary winding is represents as shown in fig. 1(b). The primary winding is divided into three coils a, b and c.
Second, the model has to be excited to set up the initial conditions. Since the sub coil with the maximum number of turns contributes more in the magnetic field, the current source is input into that sub coil with the maximum number of turns contributes more in the magnetic field, the current source is input into that sub coil and all the other coils are open-circuit. In the above example, since sub coil c in fig. 1(b) represents more turns, respectively, they are energized by the current source.
Finally, the boundary conditions for the problems have to be specified. Balloon boundary is a doped in the simulation. Balloon boundaries model the region outside the drawing space as being nearly “infinitely” large; thus, it can effectively isolate the model from other voltage sources.
To simulate an incipient internal winding fault, a model of degraded insulation before break down replaced the switch between two turns to model an incipient winding fault.
3. TLM MDELLING OF TRANSFORMER WITH INTERNAL SHORT CIRCUIT FAULTS
Large power transformers are a class of very expensive and vital components of electric power systems. Since, it is very important to minimize the frequency and duration of unwanted outages, there is a high demand imposed on power transformer protective relays; this includes the requirements of dependability associated with normal-operations, security associated with no false tripping, and operating speed associated with short-fault clearing time. Protection of large power transformers is a very challenging problem in power system relaying (Goshen et al., 2004; Mao and Agawam, 2001).
The development and the validation of algorithms for transformer protection require the preliminary determination of power transformer model. This model must simulate all the situations that will be chosen to chosen to study the behaviour of the protection algorithm. In particular, it must allow for the simulation of internal and external faults. Most of the electromagnetic transient programs available are able to accurately simulate other phenomenon occurring in the transformer like inrush magnetizing current, exciting current and external faults of a power transformer, and applying it to test a transformer protection algorithm, would provide a trust validation ( Lei fried and Fusser, 1999; 1997; Meagher, 2001).
The deteriorating insulation between the turns is a major cause of incipient internal winding faults in transformers. During the operation of the transformer, a strong electric field is applied to the dielectric material. It can result in the aging and deterioration of the insulation. The relevant factors generally recognized as causing the aging and deterioration of an insulation include thermal stresses, electrical stresses, mechanical stresses, moisture, and so on . Thermal stresses are caused by the internal heating due to current overloads plus ambient temperatures. Electrical stresses are cause by the voltage gradient in the insulation. Under normal operating conditions, high voltage gradients below the breakdown voltage do not cause detectable aging. However at elevated temperatures, the structure of a dielectric may be altered significantly during the aging process, and these changes will affect the electrical properties of the dielectric even before insulation failure occurs . The relaxation processes a dielectric undergoes, and hence the characteristics of dielectric loss, depend on the structure. As the structure of the dielectric molecules alters during aging, the dielectric characteristics and electrical properties change. Normal aging of the insulation may be manifested by the gradual reduction in it’s with stand capability over a long period of time. In addition, discharges would like to take place on the surface of the solid insulation or within enclosed voids. The gas ions driven by the electric field would hit on the wall of the insulation and react chemically with some of its surface layer molecules. Thus chemical and thermal degradation of the insulating material occurs at these microscopic sites. When a persistent discharge continues, it is called an arc discharge. This can result in the failure of the electric and a sudden breakdown under operating voltage.
To simulate the incipient internal winding faults completely, both the aging phase and the arcing phase have to be taken into account. Therefore, a combination of an insulation aging model and an arcing model was developed to model an incipient internal winding fault
5. TRANSFORMER INTERNAL INCIPIENT FAULT MODEL
The transformer internal incipient fault computer model is a combination of a two-dimensional nonlinear finite element analysis internal short circuit fault model and deteriorating insulation model consisting of an aging and an arcing component. These models are briefly reviewed in this following section. More details about these models can be found in
5.1 Transformer Non-Linear Model
The transformer two-dimensional nonlinear finite element model presented in applies finite element analysis to calculate the parameters for an equivalent circuit of a transformer with an internal short circuit fault. A soft Maxwell® software package was used to perform calculations. The transformer model can be exported as a SPICE sub-circuit and imported to the SIMPLORER® simulation environment as a black box with a set of interface terminals. These terminals are appropriately connected to the voltage source, load, and incipient fault model to simulate an incipient fault scenario.
5.2 Aging Model
The aging model is traditionally represented by an equivalent parallel RC network as shown in Fig. 1. In the circuit, Rp represents the loss part of the dielectric, which results from electronic and ionic conductivity, dipole orientation and space charge polarization, etc.; and Cp is the capacitance in the presence of the dielectric . the corresponding phasor diagram of the equivalent circuit is shown. Angle δ is defined as the loss angle, which represents the dielectric energy losses in the insulation. Tan δ is commonly known as the loss tangent or dissipation factor. The loss angle is usually very small for perfect insulation. It differentiates the losses in one dielectric material from those in the other one. Cos θ is the power factor of the dielectric. Equation expresses the relationship between circuit elements and the dissipation factor. When an AC voltage is applied, the capacitive component of the current is IC while the resistive component of the current is IR.
Under operating conditions of voltage and temperature, an insulating material may deteriorate in electric strength because of the absorption of moisture, physical changes of its surface, chemical changes in its composition, and the effects of ionization both on its surface and on the surface of internal voids. In general, the dissipation factor will be increased. With a proper understanding of the effects of aging factors, the changes in any electrical property, particularly dissipation factor, can be a measure of deterioration [7]. While the initial value of dissipation factor is important, the change in dissipation factor with aging may be much more significant.
Vw is the maximum voltage across the winding and Nw is the number of turns of the winding. Then the voltage VT was applied to the insulation sample as shown in Fig. 2. Using finite element analysis, the energy U, stored in the insulation, was calculated. Then the capacitance was calculated based on the geometry information and the material properties of the insulation using (3).
After calculating Cp, a reasonable dissipation factor representing a particular degradation level of the insulation was chosen. Since the equivalent capacitance does not change much, the resistance in the equivalent circuit can be calculated by manipulating (1) for different dissipation factors. For instance, Table I shows the calculation results for perfect insulation between adjacent turns.
According to the manufacturer specification for the type of the insulation assumed in this work, the dissipation factor for the perfect insulation is 0.006. When m turns are involved in the primary or secondary incipient fault, the equivalent circuit parameters - Rpm and Cpm- were calculated by (4) using the equivalent circuit parameters for adjacent turns.
5.3 Arcing Model
A persistent fault in a transformer would eventually involve an arc which is defined as a continuous luminous discharge of electricity across an insulating medium, usually accompanied by the partial volatilization of the electrodes the arcing characteristics can be illustrated as in Fig. 3. This is a simple case where arcing happens in a resistive load circuit. The arc ignites only after sufficient voltage is across the gap. This period is called the burning period. Arcing voltage is almost a square wave, except for the transient near current zero. In addition, arcing is a random and stochastic phenomenon. The arc extinguishes when voltage drops below that
needed to sustain the arc. This period (Δt in the figure) of effectively zero current around the virtual zero point of
the sinusoid is called the extinction period. The characteristics of an arcing current have to be considered in the simulation model to obtain a representative circuit that can take the place of the arc thereby obtaining accurate analysis results. But, of course, it is known that the arc is by no means a simple circuit element For the purpose of modelling arcing phenomena between adjacent turns of the transformer, where the arc gap is very small, the
circuit shown in Fig. 4 can be used. The voltage E is a random square wave representing the equivalent arcing voltage in the burning period. It has been shown that the arc voltage is usually flap-topped and that the magnitude of arcing fault currents ranged from 57% to 100% of the available short-circuits current. During periods of effective current zero, however, arc can be represented by a high nonlinear resistance, R (t), which increases with time. Switches S1 and S2 control the burning and extinction period of an arc, respectively. When arcing is in the burning period, S1 is closed and S2 is open. In the extinction period, S2 is closed and S1 is open. Otherwise, both switches are open.
7. SIMULATION RESULTS
Using the simulation circuit, various studies were performed, which are discussed in the following sections. Even though, most of the figures were related to secondary fault cases, primary winding faults were also simulated similarly. In fact, conceptually there are no differences between simulating primary faults and secondary winding faults. The following procedure was followed for a primary fault case. Firstly, the corresponding FE transformer non-linear model was imported and assigned to the transformer block. Secondly, the necessary changes were made to the connectivity scheme of the transformer terminals to represent a primary fault case. Lastly, the aging and arcing model parameters were tuned accordingly. Beginning with the third cycle, at time 33.33 ms, the transformer underwent an advanced insulation deterioration, which was simulated by switching the aging model on. Comparing with the previous normal condition, the primary current increased, and the secondary current and voltage decreased slightly. The primary voltage did not change much since it is biased by the Thevenin voltage of the upstream network. The amount of changes in the terminal values directly depends on the severity of the aging or the degradation level of the insulation represented by Rpm in the simulation model. The aging activity was later accompanied by the arcing current at the end of the forth cycle, at time 66.66 ms. As seen in the figure, the arcing happened randomly in the positive half cycle, negative half cycle or both. The peak primary current when arcing occurred was considerably larger than that of the primary current with
8. CONCLUSIONS
This paper presented a new transformer model to simulate an internal incipient winding fault. The new transformer model was implemented by combining deteriorating insulation model with a finite element analysis internal short circuit fault model. The new deteriorating insulation model, which includes an aging model and an arcing model, was developed based on the physical behaviour of aging insulation and the arcing phenomena occurring when the insulation was severely damaged. The aging model and the arcing model of the insulation were connected in parallel to produce a parallel combination insulation model. The parallel combination insulation model was combined with the internal short circuit model to predict the terminal voltages and currents of the transformer under various incipient fault conditions. The characteristics of the terminal currents and circulating current in the faulted winding were analyzed in time domain and frequency domain.
Some incipient-like fault cases obtained from internal short circuit field tests were also analyzed in time domain and frequency domain. The comparisons between the incipient-like fault experimental test cases and simulation results showed that terminal behaviours of the faulted transformer obtained from the field tests show some similar characteristics to those from simulations using the parallel combination model. In future work, the incipient fault transformer model will be used to generate a database of incipient internal winding faults in distribution transformers for the development of intelligent transformer fault detection techniques.
9. REFERENCES
[1] C. E. Lin, J. M. Ling, and C. L. Huang, “Expert system for transformer fault diagnosis using dissolved gas analysis,” IEEE Trans. Power Del., vol. 8, pp. 231–238, Jan. 1993.
[2] W. Yuen, H. Zhuxiao, and Z. Jinxing, “Expediting cellulose insulation aging evaluation and life prediction thorough degree of polymerization measurements,” in Proc. 3rd Int. Conf. Properties Appl. Dielectric Materials, 1988, pp. 328–331.
[3] A. Nora, K. Nakamura, T. Watanabe, and T. Morita, “Acoustic-based real-time partial discharge location in model transformer,” in Proc. ICSPAT, pp. 1077–1082.
[4] J. Bake-Jensen, B. Bake-Jensen, and S. D. Michelson, “Detection of faults and aging phenomena in transformers by transfer functions,” IEEE Trans. Power Del., vol. 10, pp. 308–314, Jan. 1995.
[5] P. Bastard, P. Bertrand, and M. Meunier, “A transformer model for winding fault studies, ”IEEE Trans. Power Del., vol. 9, pp. 690–699, Mar. 1994.
[6] A. Guzman, D. Hour, and S. E. Zocholl, “Transformer modelling as applied to differential protection,” in Proc. Can. Conf. Elect. Compote. Eng.,vol. 1, 1996, pp. 108–114.
[7] H.Wang and K. L. Butler, “Modelling transformer with internal winding faults by calculating leakage factors,” in Proc. 31st North Amer. Power Symp., San Luis Obispo, CA, Oct. 1999, pp. 176–182.
[8] H. Wang, P. Palmer-Buckle, and K. L. Butler, “Transformer models for detection of incipient internal winding faults,” in Proc. 30th North Amer. Power Symp., Cleveland, OH, Oct. 1998, pp. 90–98.
[9] H. Wang and K. L. Butler, “Finite element analysis of internal winding faults in distribution transformers,” IEEE Trans. Power Del., vol. 16, pp.422–427, July 2001.
[10] J. Hong, Y. Lee, K. Kim, H. Shin, and S. Kim, “Dielectric characteristics on insulation oil due to the aging by electron beam irradiation,” in Proc. Fifth Int. Conf. Properties Appl. Dielectric Materials, vol. 1
No comments:
Post a Comment
leave your opinion