27-04-2011, 02:56 PM
Presented by:
MANAS ANKIT DATTA
[attachment=12944]
INTRODUCTION :
Incipient fault detection in transformer can provide early warning of electric failure and could prevent catastrophic losses. To develop transformer incipient fault detection technique a transformer model to simulate internal incipient fault is required. The model for such an operation was implemented by combining deteriorating insulation model with an internal short circuit model . The internal short circuit model was developed using finite element analysis.
The deteriorating model includes an aging and an arching model connected in parallel .
The characteristics of the incipient fault from the simulation were compared with with those from some potential experimental incipient fault cases
INDEX TERMS :
1. Distribution transformer
2. Finite element analysis
3. Aging
4. Arching
5. Internal incipient winding fault
6. modeling
DISTRIBUTION TRANSFORMER
Transformer are essential and important part of power systems. Online condition monitoring of transformer can provide early warning of electric failure and avoid catastrophic losses . However the implementation of existing monitoring methods tend to cost too much when applied to distribution transformer .A study shows that 70-80% of failures are caused by failure between turns. Therefore an online fault detection method was developed for distribution transformer that utilizes terminal behaviors of transformer
To make an accurate diagnostic decision , transformer internal winding must be characterized by analyzing quantity of data which could be generated through computer simulation or field experiments.
Finite Elemental Analysis
The finite elemental method is a numerical method to obtain approximate solutions to boundary value problems of mathematical physics . It has become a very important tool to solve electromagnetic problems because of it’s ability to model geometrically and compositionally complex problems.
Using finite elemental analysis to solve problem 3 stages
1. Meshing the problem space into contiguous element of suitable property and assigning appropriate value to material parameters
2. Model has to be excited to set up initial conditions
Specify the boundary conditions
There are 2 methods to couple finite element model with circuit equations
1. Direct coupling ( circuit equations are directly incorporated into field equations and solved simultaneously)
2. Indirect coupling(circuit simulation and coupling between field and circuit model are handled separately)
Transformer model using FEA
The 2D Magneto static solver was in Maxwell package was used to compute mutual and leakage inductance and export an equivalent circuit in format of SPICE sub circuits.
The normal transformer illustrated in fig 1 was modeled in the solver.
Simulation results :
Based on simulation system in above figure , the normal case and some fault cases were simulated . For a normal transfer model the RMS value of terminal voltages and currents from the simulation and the calculated expected results were almost equal.
Aging of transformer winding :
Aging of insulation is temporal function of temperature , the quantity of humidity and oxygen .with the help of advanced manufacturing process humidity and oxygen can be reduced. Since the temperature distribution is uneven, hot spot temperatures can have devastating effects.
In considering electric behavior of dielectric material it’s convenient to follow equivalent parallel circuit as shown above.
Arching in transformer winding
An arc is defined as a continuous luminous discharge of electricity across an insulating medium usually accompanied with partial volatilization of electrodes.
The above figure shows the simple case of arching current and voltage in a resistive load circuit.
The dotted line is the arching current, solid line is arching voltage , the dash dot line is the system voltage.
Internal incipient winding fault :
S1 and S2 are time controlled switches. If there is no arching in the circuit, S1 is closed and S2 is open and the value of E is zero. When E equals E(m) the model represents arching .when S1 is closed and S2 is open arching is in burning period .when S1 is open and S2 is closed arching is in extinction period.
Parallel combination model for insulation
The above figure represents the combinational model. The parallel model represents a perfect insulation when S1 and S2 are open , the value of e is zero and R(pm) is a very large resistance so that current in the insulation is almost zero. To represent deteriorating insulation R(pm) is decreased to a small value. When the random square wave E is zero , S1 and S2 are open. The model represents the non arching deteriorating insulation . When the value of e is E(m) , the model represents degrading insulation with arching. In this case when S1 is closed and S2 is open, the model simulates the burning period. If S1 is open and S2 is closed it models the extinction period.