Chaitanya Ashok Baone

Fast contingency screening and ranking for small signal stability assessment

Utilities today are in need of tools and techniques that will enable them to predict the dynamic stability and reliability of the grid in the real-time. The problem is challenging because of the large number of contingencies that are to be simulated. In this paper a fast method for power system contingency screening and ranking for small signal stability assessment is presented which essentially reduces the number of contingencies for detailed evaluation. The proposed method avoids repeated computation of eigenvalues for all possible post-contingency scenarios. Instead, the eigenvalues corresponding to critical modes for post-outage conditions are estimated based on first-order eigenvalue-sensitivity using just the nominal condition eigenvalues and post outage system state matrices. Since a critical outage condition can produce a large change in the eigenvalues, the first order prediction might not have acceptable accuracy. To overcome this issue, a second-order correction is applied which needs the computation of the eigenvectors corresponding to the eigenvalues of interest. A significant reduction in computation time for evaluation of post contingency eigenvalues using the proposed method is demonstrated. A case study on a 16-machine NETS-NYPS system shows promising results.

Measurement based static load model identification

Accurate load models are necessary to enable multiple power system planning and operation applications. The current day utility practice relies on model parameters that are based on old survey results/tests and often do not accurately capture the behavior of loads across different operating conditions. This results in conservative estimation of loading margin, leading to under-utilization of the assets. Static load models are typically chosen as purely constant power models, resulting in lack of understanding of the voltage dependence of loads. This paper considers the problem of determining the voltage sensitivity of aggregate loads and its impact on voltage stability. The role of voltage sensitivity of loads in determining accurate PV curves is demonstrated. Additionally, a novel recursive least mean squares based algorithm is proposed to estimate a generalized static load model that not only estimates the model parameters but also captures typical patterns in load variation with respect to time, season, temperature, etc., by utilizing measurement data from switching events. The proposed approach eliminates the need to rely on stage tests to obtain the data samples, and instead utilizes available measurements from events recorded over time. The efficacy of the proposed approach is demonstrated using actual field data.

Optimal day-ahead scheduling for microgrid participation in frequency regulation markets

As microgrid installations are steadily growing in the United States and around the world, widespread adoption of commercial microgrids would rely upon the economic benefit to the owners and operators. With the introduction of new market mechanisms and growing penetration of non-traditional generation assets, there is an increasing need and interest in allowing distributed assets to participate in traditional grid services such as frequency regulation. This paper considers the problem of determining the optimal balance of energy and ancillary services for individual microgrid generation assets to participate in such markets. An optimization framework that maximizes the predicted performance of the microgrid over a day-ahead time horizon while accounting for individual asset constraints is proposed. Simulation results on a realistic test system with practical considerations are presented.

Chapter 28 – Control of Power Systems with High Penetration Variable Generation

This work will consider new approaches to optimal control design and distributed estimation, to allow renewable resources and storage technologies to more effectively contribute to system-wide goals of grid frequency regulation and electromechanical stability enhancement. Examining a specific case study of wind turbine generation and battery energy storage, this work will illustrate how optimal design can best exploit the very different dynamic characteristics of these disparate types of resources, while respecting the control saturation limits in which they much operate (e.g., power and stored energy limits for a battery, response bandwidth limits for wind turbine blade pitch control). Coupled with this will be techniques for local observation of the overall grid’s dynamic state (made tractable by limiting attention to a low dimension subspace), allowing distributed resources to contribute to system wide control objectives.

A modal decomposition approach to distributed observation and control for grid integration of renewable generation

To enable renewable generation to fully contribute to reliable grid operation with higher levels of penetration, it is important that such resources contribute to grid control, including faster time scale active power control that is valuable to small-signal electromechanical stability. However, given the small size and distributed nature of many renewables, it is unrealistic to assume that they will employ the type of centrally coordinated control schemes typical of large, central station synchronous generators. This paper proposes a distributed control design approach that allows each distributed renewable generator or energy storage device to be responsible for improving damping of a single oscillatory electromechanical mode. The consequence is that one may employ a static state feedback design (a classic LQ design is employed here) that requires information of state behavior only within a two-dimension invariant subspace associated with the mode of interest. Using a modal-focused Hautus matrix test of obervability, we develop simple tests to establish grid buses from which a desired mode may be most effectively observed from either local frequency measurement, or from a local measurement plus a single remote phasor measurement unit (PMU) signal. In this framework, the state information required may then be recovered by a Luenberger observer of as little as second order. That is, the local, distributed controller on a given renewable generator or energy storage device uses only one or two measurements, may be as simple as second order, and yet contributes to the system objective of improving modal damping.

Control of Power Systems with High Penetration Variable Generation

This work will consider new approaches to optimal control design and distributed estimation, to allow renewable resources and storage technologies to more effectively contribute to system-wide goals of grid frequency regulation and electromechanical stability enhancement. Examining a specific case study of wind turbine generation and battery energy storage, this work will illustrate how optimal design can best exploit the very different dynamic characteristics of these disparate types of resources, while respecting the control saturation limits in which they much operate (e.g., power and stored energy limits for a battery, response bandwidth limits for wind turbine blade pitch control). Coupled with this will be techniques for local observation of the overall grid’s dynamic state (made tractable by limiting attention to a low dimension subspace), allowing distributed resources to contribute to system wide control objectives.