The weighted least squares (WLS) estimator is commonly employed to solve the state estimation problem in today's power systems, which are primarily measured by SCADA measurements, including bus power injection, branch power flow and bus voltage magnitude measurements. Despite being widely used, WLS estimator remains to be non-robust, i.e., it fails in the presence of bad measurements. The so called least absolute value (LAV) estimator is more robust, but is not widely used due to its higher computational cost. LAV estimator has the desirable property of automatic bad-data rejection provided that the measurement set does not include any “leverage measurements”. There will be two important advantages to using LAV estimator when systems are observed by phasor measurements: 1) if the measurement set consists of only phasors, the leveraging effect of measurements can be easily eliminated by strategic scaling; 2) computational performance of LAV estimator will become competitive with that of WLS due to the linearity of the estimation problem for phasor measurements. This paper demonstrates these advantages and argues that the LAV estimator will be a statistically robust and computationally competitive estimator for those power systems that will be measured entirely by PMUs.

This paper is concerned about effective incorporation of a limited number of phasor measurement unit (PMU) measurements into the state estimation solution in order to enhance the state tracking accuracy and speed. This is particularly important for fast changing system conditions when rapid control action may be necessary based on the onset of voltage instability due to an unexpected event. Under such conditions, SCADA-based state estimators may be too slow in capturing the changes in system state due to the low scan rates of the order of seconds. PMU measurements on the other hand provide refresh rates of milliseconds; however, they may not provide full system observability. In this work, a solution is proposed in order to address the lack of observability by the PMU measurements while maintaining fast tracking capability as well as estimation robustness against pseudo-measurement and PMU errors. Robustness is accomplished via the use of the so-called least absolute value estimator which minimizes L1 norm of the residuals. Simulation results are provided to illustrate the expected benefits of this implementation on test systems.

Power systems do not always operate under fully balanced conditions because of unbalanced loads and untransposed transmission lines. Under such conditions, commonly used positive sequence equivalent will no longer be valid. On the other hand, use of full three-phase model will add significant computational burden for the considered application. This paper describes a simple approach to address this issue for the specific application of state estimation. The main idea is to exploit the linearity of the estimation problem when the system is measured solely by phasor measurements. The linear problem lends itself to decoupling via modal transformation unlike the nonlinear formulation with SCADA measurements. A further improvement is introduced via the use of the least absolute value (LAV) estimator, which has the desirable property of automatic bad-data rejection. The paper illustrates how a robust LAV estimator designed for a positive sequence model can be efficiently used for the solution of an unbalanced three-phase state estimation problem.

Given the rapid increase in the number of phasor measurement units (PMU) installed in substations in recent years, it is not unrealistic to assume that most transmission systems will be fully observable by only PMU measurements in the not so distant future. Such a scenario will enable faster and numerically easier state estimation solution which will be non-iterative due to the linearity of the PMU measurement equations. However, as will be shown in this paper, the benefits do not end there and a closer look at these equations reveals further possibilities of simplification and therefore improved performance of state estimators that use only PMU measurements. Specifically, the paper formulates an alternative decoupled solution, which reduces the size of the state estimation problem, increases computational performance yet maintains high solution accuracy. Both state estimation as well as bad data processing can be efficiently carried out within the developed decoupled formulation.

A new electric arc furnace (EAF)-specific model based on field measurements of instantaneous furnace voltages and currents has been proposed. This model presents the dynamic behavior of the EAF system, including all parts, i.e., the EAF transformer, the secondary circuit, the electrode movements, and the arc. It consists of a cascade connected variable-resistance and -inductance combination to represent the time variation at the fundamental frequency and a current source in parallel with it to inject the harmonics and interharmonics of the EAF current. The model takes into account several typical tap-to-tap periods of the specific EAF operation. This model is particularly advantageous for power quality (PQ) analysis and development of flexible alternating current transmission system solutions for PQ problem mitigation of a given busbar supplying single- or multi-EAF installations. The validity of the proposed model has been verified by comparing EMTDC/Power Systems Computer-Aided Design simulations of the model with the field measurements. The results obtained have shown quite satisfactory correlation between the proposed model and the actual EAF operation.

Current state estimators employ the Weighted Least Squares (WLS) estimator to solve the state estimation problem. Once the state estimates are obtained, Chi-Square test is commonly used to detect the presence of bad data in the measurement sets. Regretfully, this test is not entirely reliable, that is, bad data existing in the measurement set could be missed for certain cases. One reason for this is the approximations used to compute the bad data suspicion threshold, which is set based on an assumed chi-squares distribution for the objective function. In this paper, a modified metric is proposed in order to improve the bad data detection accuracy of the commonly used chi-square test. The bad data detection performance of the proposed test is compared with that of conventional chi-square test.