track of power system impedence paramete

Views:
 
     
 

Presentation Description

No description available.

Comments

Presentation Transcript

Slide 1: 

15 March 2010 1 Presented By: K. Santosh Kumar Patro Regd No.:- 0903019 V S S U T Online Tracking of Power System Impedance Parameters and Field Experiences

Slide 2: 

15 March 2010 2 Presented By: K. Santosh Kumar Patro Regd No.:- 0903019 V S S U T Why we calculate impedance parameter ? Use to calculate Short circuit currents and to verify models of power networks. Also needed for VAr compensator and harmonic filter design to avoid creating resonance condition. In recent years, the equivalent impedance has been used as parameter for fault and protection calculation. Application in power electronics device for limiting voltage waveform distortion. To estimate voltage stability margin and maximum loadability of the system.

Slide 3: 

15 March 2010 3 Presented By: K. Santosh Kumar Patro Regd No.:- 0903019 V S S U T PROPOSED METHODS: Two types methods Invasive Non-Invasive Least Squared Methods Signal Processing Based Method Other Methods like Voltage Instability Predictor (VIP), resistance sign based methods

Slide 4: 

15 March 2010 4 Presented By: K. Santosh Kumar Patro Regd No.:- 0903019 V S S U T This paper proposes a new algorithm which can be implemented into the power monitors widely available in load serving substation. It is independent of the load models and does not require synchronized data. Proposed Algorithms for Impedance Estimation methods Three-point Method Multipoint Method

Slide 5: 

15 March 2010 5 Presented By: K. Santosh Kumar Patro Regd No.:- 0903019 V S S U T Three Point Method At time t1 At time t2 At time t3

Slide 6: 

15 March 2010 6 Presented By: K. Santosh Kumar Patro Regd No.:- 0903019 V S S U T Multipoint Method Very sensitive to noise and transient in the voltage and current measurement For n set of Measurement εxi and εyi are estimation error

Slide 7: 

15 March 2010 7 Presented By: K. Santosh Kumar Patro Regd No.:- 0903019 V S S U T The vector Z that minimizes the function f(Z) will be the solution to the problem. The minimization can be done through iterative based methods. In this the Gauss- Newton method has been used. The goal is to minimize error in a certain no of measurement and we can define the function and minimize it:

Slide 8: 

15 March 2010 8 Presented By: K. Santosh Kumar Patro Regd No.:- 0903019 V S S U T In order to compare the accuracy of the three-point and the multipoint algorithm, some case studies have been done to investigate the sensitivity of the two algorithms to the measurement noise. For this purpose, several simulation case studies have been prepared and different random noise levels have been added to the measured voltages and currents. By using four different numbers of points have been applied to calculate the system parameters. To calculate each plotted value in Fig. , 300 cases have been run and the estimation errors have been calculated for each case and then the average has been plotted. The error is calculated as follows:

Slide 9: 

15 March 2010 9 Presented By: K. Santosh Kumar Patro Regd No.:- 0903019 V S S U T In Fig. the accuracy of calculated Rs and Xs is plotted as a function of SNR for the three-point and multipoint algorithms. The simulation results in Fig. show that increasing the number of points used for calculations reduces the sensitivity of the algorithm to the measurement noise. This assumption is valid while all the system side parameters remain constant.

Slide 10: 

15 March 2010 10 Presented By: K. Santosh Kumar Patro Regd No.:- 0903019 V S S U T Practical Consideration: Disturbance Side Detection: The proposed three-point algorithm can certainly determine the disturbance side for the measurements. Defining Vi = Vxi + Vyi taking the square of both sides of (5), we can rewrite them as follows: Using the three equations, we can eliminate Es from the equations and calculate Xs as a function of Rs, so that

Slide 11: 

15 March 2010 11 Presented By: K. Santosh Kumar Patro Regd No.:- 0903019 V S S U T Equation (15) provides a simple second-order equation for that can easily be solved. It has been derived based on the assumption that the system side parameters remain unchanged for three different system conditions and if for any reason the system side parameters vary for the three points, (15) is meaningless and will not have any solutions. The only condition for (15) to have a real solution is Δ>=0(Δ = b2 - 4ac); therefore the sign of Δ can be used as an index for determination of disturbance side. If Δ is greater than or equal to zero the necessary condition which is unchanged system side parameters, should be correct and we will have one or two solutions. Negative Δ can happen due to the variation of system side parameters, measurement noise, and load switching transients that sometimes exist in power system measurements. And finally

Slide 12: 

15 March 2010 12 Presented By: K. Santosh Kumar Patro Regd No.:- 0903019 V S S U T After calculating Rs, we can calculate both Xs and Es from (6) and then the feasibility of solutions can be checked by using the following rules: Rs, Xs and Es should always be positive; X /R ratio should have feasible values. Therefore, the proposed three-point algorithm can be used for selecting the data that can be used for the multipoint algorithm. If (15) has solution for every three consecutive points selected from a set of measured data, we can conclude the system side parameters did not have variation for that set of measurement. In this case instead of three points we can use more measurement points for estimation of system impedance parameters which improves the accuracy of the algorithm.

Slide 13: 

Load Fluctuation Index: The proposed algorithm uses the fluctuations of loads connected to the system to calculate the system parameters. Therefore one of the conditions for the input data of the algorithm is variation of active and reactive power of the load at the substation bus. Since mathematically, variation of one of the variables, P or Q ,would be enough for the equations to be solved, the fluctuations can be defined as the summation of absolute variation of and , as stated in (16) 15 March 2010 13 Presented By: K. Santosh Kumar Patro Regd No.:- 0903019 V S S U T

Slide 14: 

15 March 2010 14 Presented By: K. Santosh Kumar Patro Regd No.:- 0903019 V S S U T THE PROPOSED ALGORITHM FOR THREE-PHASE SYSTEMS: The power systems are usually designed as three phase systems. Assuming that the system is balanced, the three phase voltages and currents can be transformed to positive, negative and zero sequence voltages and currents. Since the sequences are treated as single phase circuits, the proposed algorithms can be used to calculate the positive, negative, and zero sequence equivalent circuit parameters of the system. The proposed algorithm for three phase systems is explained in the following steps. Three to five seconds waveforms of three phase voltages and currents of the feeder is captured with 15.36-kHz sampling rate. Each cycle of the 60-Hz three-phase voltage and current waveforms are transformed to frequency domain using Fourier transformation. The positive sequence voltages and currents for each cycle of the system frequency are calculated using sequence transformation matrix. The multipoint algorithm, considering the practical issues, is applied to calculate the positive sequence equivalent circuit parameters of the system.

Slide 15: 

15 March 2010 15 Presented By: K. Santosh Kumar Patro Regd No.:- 0903019 V S S U T

Slide 16: 

15 March 2010 16 Presented By: K. Santosh Kumar Patro Regd No.:- 0903019 V S S U T

Slide 17: 

15 March 2010 17 Presented By: K. Santosh Kumar Patro Regd No.:- 0903019 V S S U T Simulation and Experimental Verification: Simulation result:-

Slide 18: 

15 March 2010 18 Presented By: K. Santosh Kumar Patro Regd No.:- 0903019 V S S U T Experimental Results:-

Slide 19: 

15 March 2010 19 Presented By: K. Santosh Kumar Patro Regd No.:- 0903019 V S S U T Implementation and Field test Verification: Field measurement were conducted on August and September 2007 at load serving station in Alberta, Canada. The data processing was performed in power lab at university of Alberta to determine the impedance parameters of the system. Instrument Set Up: NI-6020E 12-bit DAS with 15.36 kHz sampling rate controlled by laptop for recording Captured waveforms are 3-φ voltage and current at the point of metering in load serving substations. CTs and PTs have been used to step down current and voltage respectively too measurable data. Measurement are taken in every minute a 5 second of data window of 3-φ voltage and current waveforms are captured.

Slide 20: 

15 March 2010 20 Presented By: K. Santosh Kumar Patro Regd No.:- 0903019 V S S U T Substation 1:-

Slide 21: 

15 March 2010 21 Presented By: K. Santosh Kumar Patro Regd No.:- 0903019 V S S U T

Slide 22: 

15 March 2010 22 Presented By: K. Santosh Kumar Patro Regd No.:- 0903019 V S S U T

Slide 23: 

15 March 2010 23 Presented By: K. Santosh Kumar Patro Regd No.:- 0903019 V S S U T Substation 2:-

Slide 24: 

15 March 2010 24 Presented By: K. Santosh Kumar Patro Regd No.:- 0903019 V S S U T

Slide 25: 

15 March 2010 25 Presented By: K. Santosh Kumar Patro Regd No.:- 0903019 V S S U T

Slide 26: 

15 March 2010 26 Presented By: K. Santosh Kumar Patro Regd No.:- 0903019 V S S U T CONCLUSION The following conclusions are drawn. •) Three to five seconds’ data window with sufficient variations is enough for the estimation of impedance parameters. •) The 256 sample per cycle rate of data acquisition system is sufficient for data collection process. •) The algorithm does not depend on the load model and works with natural variations of any kind of loads. •) Synchronized measurements are not required for calculations, and the method is not sensitive to changes of system frequency and harmonics. •) The required information can be obtained from the meters that are already installed at the substation buses. The proposed algorithm has been verified using simulations, experiments and real field tdata. The verification results show that the proposed method can be implemented for online monitoring of power system impedance parameters.

Slide 27: 

15 March 2010 27 Presented By: K. Santosh Kumar Patro Regd No.:- 0903019 V S S U T Thank u

authorStream Live Help