Liisa Haarla

Comparison of Three Electromechanical Oscillation Damping Estimation Methods

This paper describes three data driven methods to monitor electromechanical oscillations in interconnected power system operation. The objective is to compare and contrast the performance of the methods. The accuracy of damping ratio and frequency of oscillations are the measures of the performance of the algorithms. The advantages and disadvantages of various techniques and their limitations to measurement noise have been considered while assessing performance. The target frequency and damping are computed using the Nordic power system simulation model.

Performance of wavelet-based damping estimation method under ambient conditions of the power system

The paper discusses performance evaluation of damping estimation methods and specifically evaluates the performance of a wavelet-based damping estimation method. The focus is on the damping estimation under the ambient conditions of the power system. Damping estimation results are reviewed in case of the simulated data because then real damping is known and can be compared with the estimated damping. Various degrees of complexity of the simulation models are used: linear single and double pole pair models, and a detailed nonlinear system model of the Nordic power system. In the linear models, the real damping of the modes can be analytically calculated and compared with the estimated damping. Damping estimation performance is studied in different operating conditions of the studied simulation models. The main results are: frequency estimates are better than the damping estimates, other oscillation modes may affect the damping estimation, poor damping can be estimated more accurately than high damping, and the performance of the method is nearly similar regardless of the complexity of the simulation model.

A Two-Level Probabilistic Risk Assessment of Cascading Outages

Cascading outages in power systems can lead to major power disruptions and blackouts and involve a large number of different mechanisms. The typical development of a cascading outage can be split in two phases with different dominant cascading mechanisms. As a power system is usually operated in N-1 security, an initiating contingency cannot entail a fast collapse of the grid. However, it can trigger a thermal transient, increasing significantly the likelihood of additional contingencies, in a “slow cascade.” The loss of additional elements can then trigger an electrical instability. This is the origin of the subsequent “fast cascade,” where a rapid succession of events can lead to a major power disruption. Several models of probabilistic simulations exist, but they tend to focus either on the slow cascade or on the fast cascade, according to mechanisms considered, and rarely on both. We propose in this paper a decomposition of the analysis in two levels, able to combine probabilistic simulations for the slow and the fast cascades. These two levels correspond to these two typical phases of a cascading outage. Models are developed for each of these phases. A simplification of the overall methodology is applied to two test systems to illustrate the concept.

Statistical analysis of large scale wind power generation using Monte Carlo Simulations

As more wind power generation is installed, the effect of wind power on the electric power system is becoming increasingly important. This paper presents two time series models that can be used in Monte Carlo simulations to assess the risk of very high or low wind speeds occurring contemporaneously in multiple locations. The suitability of the models is assessed for existing measured locations and new non-measured locations. The simulation results are verified against measurements from 19 locations from Finland. Also, an example scenario is given to show the effect of geographical spread on the aggregate power generation of multiple wind power generation units.

Modal analysis of power systems with eigendecomposition of multivariate autoregressive models

Analysis of electromechanical modes provides substantial information regarding the power system stability. This paper introduces a novel approach to the measurement based modal analysis of power systems by using a multivariate autoregressive model (MAR). The MAR model utilizes data that are simultaneously measured from several locations in the power system through a wide area monitoring system (WAMS). The performance of the MAR model is analyzed by applying it to data generated with a 39-bus New England test system. In addition, the model is utilized for analyzing data generated with a detailed simulation model of the Nordic power system. The results indicate that the frequencies and damping ratios of electromechanical oscillatory modes can be accurately analyzed with the eigendecomposition of the MAR model. Thus, the MAR model is a promising identification technique for wide-area monitoring of electromechanical oscillations.

Analysis of electromechanical modes using multichannel Yule-Walker estimation of a multivariate autoregressive model

Stability of power systems can be analyzed by monitoring the electromechanical oscillatory modes. This paper investigates the use of multichannel Yule-Walker (YW) estimation of a multivariate autoregressive model (MAR) for the measurement based modal analysis of power systems. The proposed YW-MAR method utilizes data that are simultaneously measured through a wide area monitoring system (WAMS). The performance of the method is analyzed by applying it to data generated with the New England test system. The results indicate that the frequencies and the damping ratios of electromechanical oscillatory modes can be accurately analyzed by using the YW-MAR method and the method is not significantly affected by measurement noise or losing a measurement signal.

Recent developments in modal estimation of power system electromechanical oscillations

This paper presents three data-driven methods to monitor electromechanical oscillations in power systems during ambient operation. The characteristics of the methods are compared and the differences in the characteristics are explained. The results indicate that all three methods are suitable for estimating the damping and frequency of electromechanical oscillations. The paper also studies the validity of the assumption that the load variations in power systems are Gaussian distributed. This is a common assumption in damping estimation methods. The assumption is shown to be valid for the studied loads.

Robustness and performance of a multi-objective Model Predictive Static Var Compensator controller

This paper presents the robustness and performance study of a wide-area multi-objective optimization based Model Predictive Control (MPC) strategy for Static Var Compensator (SVC). Many MPC control strategies presented in the literature have been based on single-objective optimization and focused merely on damping of the electromechanical oscillations. In this paper, the MPC controller is designed both for the oscillation damping and voltage control using multi-objective optimization approach. With this approach, the controller can improve the system stability in various operating conditions where either the damping or voltage stability is more important. The performance of the proposed MPC controller is compared with a local SVC controller using nonlinear time domain simulations. A simplified model of the power system in the Northern parts of Finland and Norway is used. The conducted study indicated that the proposed MPC controller could improve the system stability in varying operating conditions.

Multi-objective Model Predictive Control for damping inter-area power oscillations

This paper introduces a multi-objective optimization based Model Predictive Control (MPC) strategy for damping inter-area oscillations in power systems. The previous MPC control strategies presented in the power system literature have been based on single-objective optimization, but this paper extends the approach to multi-objective optimization. The controller is designed using a linear discrete time state space model and calculates control signals in centralized fashion for local control devices. The optimal control sequence is calculated as a solution to a multi-objective optimization problem with two objective functions. The controller is combined with a Kalman filter state estimator using voltage angle measurements. The performance of the proposed multi-objective optimization based MPC controller is compared against a single-objective optimization based MPC controller using New England New York test system model.

Statistical Analysis and Forecasting of Damping in the Nordic Power System

This paper presents an application of multiple linear regression (MLR) to extract significant correlations between damping of electromechanical modes and system operating conditions and to forecast future damping values, based on existing day-ahead market forecasts for power flows and generation. The presented analysis uses measurements from the Nordic power system. First, a static MLR model is developed to explain the variability of the damping of the 0.35-Hz inter-area mode in the Nordic system. Together with the static model, a dynamic MLR model is used for forecasting the damping 24 hours ahead, using day-ahead market forecasts. Test results indicate the proposed methods are able to correctly predict about 90% of the low damped operating conditions observed during a year, if day-ahead market forecasts are accurate. These results suggest that the methods could be used to issue early warnings about future operating conditions with low damping.