Showing papers by "Jovica V. Milanovic published in 2016"

Online Identification of Power System Dynamic Signature Using PMU Measurements and Data Mining

Abstract: This paper proposes a two-stage methodology for online identification of power system dynamic signature using phasor measurement unit (PMU) measurements and data mining. Only transient stability status is usually predicted in the literature to assist with corrective control, without the dynamic behavior of generators in the event of instability. This paper uses traditional binary classification to identify transient stability in the first stage, and then develops a novel methodology to predict the nature of unstable dynamic behavior in the second stage. The method firstly applies hierarchical clustering to define patterns of unstable dynamic behavior of generators, and then applies different multiclass classification techniques, including decision tree, ensemble decision tree and multiclass support vector machine to identify characterized unstable responses. The proposed methodology is demonstrated on a multi-area transmission test system. High prediction accuracy at both stages of identification is demonstrated.

Efficient Estimation of the Probability of Small-Disturbance Instability of Large Uncertain Power Systems

Abstract: This paper proposes a methodology to efficiently estimate the probability of small-disturbance rotor angle instability of uncertain power systems. Traditional Monte Carlo (MC) approaches are computationally intensive and inefficient, particularly when used to study low probability conditions which result in small disturbance instabilities and develop into serious outage events high impact. The proposed methodology uses importance sampling to focus on conditions which contain the high information content required to make relevant decisions about low probability events. Latin hypercube sampling (LHS) is used to efficiently bound the search space and identify operating conditions leading to a marginally stable or unstable system response. The proposed approach is demonstrated on a model of a multi-area transmission network with a significant capacity of intermittent generation connected through a multi-terminal voltage source converter-based high voltage direct current (VSC-HVDC) grid. It is demonstrated that the methodology yields accurate results with just a small fraction of the sample points required using a conventional numerical MC approach.

Assessing the Applicability of Uncertainty Importance Measures for Power System Studies

Abstract: This paper critically evaluates a number of uncertainty importance measures for use in power system stability studies. Sensitivity analysis of uncertain system parameters is vital as new technologies proliferate and the total level of system uncertainty grows. Accurate assessment of the importance of different uncertainties can guide power system operators towards parameters which will require the greatest levels of mitigation or increased monitoring in order to reduce the uncertainty and its subsequent impact. Local and global sensitivity analysis techniques are described and evaluated within this paper, including nonparametric methods, variance-based approaches, and distribution-based techniques. The techniques are illustrated using a large 295-bus realistic network model of a generic distribution system. Numerical experiments on dynamic models are used in order to assess the impact of uncertainties on the mitigation of system frequency excursions using single-site and distributed energy storage devices.

The Effect of VSC-HVDC Control on AC System Electromechanical Oscillations and DC System Dynamics

Abstract: In order to strengthen the onshore transmission network in many parts of the world, voltage-source current–high-voltage direct current (VSC–HVDC) will increasingly be utilized. The effect of the operating point of a VSC-HVDC link and the control strategies employed can substantially affect the electromechanical oscillatory behavior of the ac network as well as the dc-side dynamics. In order for the full, flexible capability of VSC-HVDC to be exploited, further study of the effects of these controllers and their interactions with ac system responses is necessary. This paper addresses this gap. Both modal analysis and transient stability analysis are used to highlight tradeoffs between candidate VSC-HVDC power controllers and to study the electromechanical performance of the integrated ac/dc model. Tests are carried out on both a generic two-area model and a large-scale realistic network with detailed ac generator and HVDC models.

Optimal Compensation of Transmission Lines Based on Minimisation of the Risk of Subsynchronous Resonance

Abstract: The paper proposes a framework for the selection of optimal configuration of line compensation devices, thyristor controlled series capacitors (TCSC) and conventional fixed series capacitors, for reducing the risk of subsynchronous resonance (SSR) in the network while maintaining required power transfer. The methodology developed in the paper is based on the robust risk evaluation of SSR that takes into consideration the severity of subsynchronous resonance and probability of its occurrence. The subsynchronous resonance risk index developed previously is used to assess the severity of subsynchronous resonance. The line outage model is employed to determine the probability of contingencies potentially leading to subsynchronous resonance. It is demonstrated that with relatively small participation of TCSC, even with the most basic control, in series compensation of lines, the risk of SSR can be successfully managed.

Study of frequency response in power system with renewable generation and energy storage

Abstract: This paper presents the first comparative analysis of frequency support provided by a large-scale bulk energy storage system (ESS) against distributed ESS in a large meshed network. In IEEE 16 machine network, 30% of synchronous generation is replaced by renewables (RES). The analysis is further extended to investigate the effect of increased penetration of RES and reduction in inertia by decreasing synchronous generation at low network loading conditions for both types of ESS arrangements.

Day-Ahead Prediction and Shaping of Dynamic Response of Demand at Bulk Supply Points

Abstract: Prediction and shaping of dynamic response of the demand will enable advanced control algorithms for active demand management as well as improved stability assessment of the power system. Based on previously developed demand disaggregation approach, this paper develops a methodology for day-ahead prediction and shaping of dynamic response of the demand at bulk supply points without having to perform field measurements. The methodology is broadly based on application of artificial neural network and Monte Carlo simulations and incorporates multiple approaches namely, load forecasting, load disaggregation and the component-based load modelling approach. The input data include standard rms measurements at bulk supply points and actual and day-ahead forecasted weather data and does not rely on having access to detailed customer surveys or high-resolution load signatures. Measured dynamic responses of the demand from the substations of the local utility are used for validation. Load shifting and shaping of dynamic response of the demand are also illustrated.

A framework to assess the effect of reduction in inertia on system frequency response

Abstract: This paper presents a framework to analyze the impact of different penetration levels of renewable generation sources (RES) and consequent reduction in inertia on grid frequency. The developed methodology is demonstrated using three operating conditions of the network. For each operating condition, the decrease in the network loading is balanced by disconnecting a part of synchronous generation. To establish the critical penetration levels of renewables and inertia limits for the grid frequency, the uncertainty of loads, intermittent and stochastic patterns of RES generation around each operating condition are simulated. The results clearly identify the critical penetration levels of RES and reduction in inertia limits of the system for frequency stability. In addition, the performed analysis quantifies the effect of primary frequency response and reduction in inertia on frequency nadir. The proposed framework is applied to the modified 16 machine and 68 bus network.

Comparison of dynamic models of battery energy storage for frequency regulation in power system

Abstract: The paper investigates the use of frequently discussed battery energy storage system (BESS) models for frequency regulation studies. Integration of a large number of renewable generation sources results in increased uncertainty in electric power generation, requiring, among the others, more frequency regulation services than before. The battery energy storage system models are compared and evaluated to assess their suitability for frequency regulation studies. The accuracy and complexity of BES models reported in the past are also discussed.

Efficient identification of critical parameters affecting the small-disturbance stability of power systems with variable uncertainty

Abstract: This paper implements an efficient sensitivity analysis (SA) technique to identify and rank critically important uncertain parameters that affect the small-disturbance stability of a power system. Identification and ranking of uncertain parameters are vital in modern power system operation due to the adoption of deregulated market structure and integration of intermittent energy resources and new types of loads. Ranking of critical uncertain parameters will facilitate better operation and control with less monitoring (targeted only on the parameters of interest) by system operators and stakeholders. The Morris screening method of sensitivity analysis has been described and implemented in this paper as the most suitable for this study based on comparison with various local and global techniques which highlighted the their comparative computational complexities and simulation time requirements. All methods have been illustrated using a modified version of the 68 bus NET-SNYPS test system. Illustrative results are provided considering varying levels of parameter uncertainties in order to establish not only the impact of system variability on parameter ranking, but also the robustness of the presented technique.

Quantification of Uncertainty in End-of-Life Failure Models of Power Transformers for Transmission Systems Reliability Studies

Abstract: Power system components have a relatively long life span and hence there is not enough data to derive an accurate end-of-life failure model. This contributes to the uncertainty in system reliability assessment. This paper discusses the quantification of the effect of uncertainty in end-of-life failure models on system reliability indices. This paper characterizes the uncertainty by a mixed aleatory–epistemic uncertainty model, where the aleatory uncertainty originated from the variability of failure events and the epistemic uncertainty originated from a lack of data. The mixed aleatory–epistemic uncertainty was propagated using two methods: Second Order Probability and Dempster–Shafer Evidence Theory (DSET). Power transformers were chosen as the case study equipment and the methods were applied on a realistic transmission network.

Pathway to cost-efficient state estimation of future distribution networks

Abstract: This paper provides a detailed overview of the techniques and applications related to distribution system state estimation (DSSE), together with the classification of various types of state estimation. The paper also provides the state-of-art techniques applied in DSSE including forecasted-aided state estimation, close-loop DSSE methods, the application of computation intelligence in DSSE and the use of smart meters and phasor measurement units in load estimation. As more and more active elements and functionalities will be integrated in future distribution network, e.g. demand-side management, the increased penetration of decentralized generation and dynamically controlled storage/devices, DSSE will be one of the critical functionalities for secure operation of future distribution networks. As a pathway to cost-efficient DSSE for future distribution networks, this paper addresses the characteristics of future distribution grids that will affect DSSE, and also discusses the techniques and data that can be used to better understand/model the network behaviors and load profiles in order to ultimately improve the accuracy of DSSE.

The influence of different storage technologies on large power system frequency response

Abstract: This paper investigates the contribution of energy storage systems to the system frequency response in a large meshed transmission system. The case studies compare the contribution of two types of energy storage technologies, i.e., battery energy storage system (BESS) and flywheel energy storage system (FES), to improve frequency response in the test network. Case studies examine the effect of different type of disturbance on frequency response in the presence of each type of storage technology. The effect of one bulk energy storage and distributed energy storage systems of the same capacity on frequency response of the power system is also investigated.

Optimisation framework for development of cost-effective monitoring in distribution networks

Abstract: This study presents a novel optimisation methodology, optimal placement of monitors (OPMPower), for optimal device/monitor placement in distribution networks. OPMPower is developed based on gradient search and particle swarm optimisation. The proposed method integrates network topology into search process via spanning trees and uses the historical experience for search guidance. The method is particularly suited for optimal placement problems in power systems. The application is illustrated on the problem of optimal monitor placement for estimation of voltage unbalance in a section of existing UK distribution network and in a generic distribution network. It is demonstrated that the proposed methodology outperforms generic integer optimisation algorithms which are widely used for optimal placement problems in the literature, for example genetic algorithms.

Probabilistic assessment of transient stability in reduced inertia systems

Abstract: A probabilistic approach to assess the transient stability of power systems with increased penetration of wind and Photo-Voltaic generation is presented in this paper. The impact on transient stability due to the intermittent behavior of Distributed Energy Resources (DERs) as well as due to their dynamic response when a disturbance happens is investigated. Moreover, the effect of conventional generation disconnection and consequently inertia reduction is studied. Apart from calculating transient stability related indices, a clustering technique is also applied to provide more information considering the impact of DERs and conventional generation disconnection on transient stability.

Assessment of Power quality performance in distribution networks part I - Measurement campaign and initial analysis

Abstract: Power quality (PQ) is a raising concern in distribution grids of modern industrialized countries. The PQ monitoring activities of distribution system operators (DSO) and consequently the amount of PQ measurement data increases continuously. To keep the routine assessment of PQ levels efficient, new and automated tools for validating, analysing and visualizing these data are required. The paper illustrates the challenges of processing long-term PQ monitoring campaigns by using measurement data at 8 different sites in public low voltage (LV) networks for more than one year (64 weeks). The paper is divided into two parts. The first part describes the measurement campaign, the pre-processing of the data and an easy and flexible way for routine compliance assessment. The second part is dedicated to the comparison of two different global PQ indices for assessing an average site performance as basis for e.g. benchmarking purposes.

Assessment of Power Quality performance in distribution networks part II - Performance Indices and ranking of network buses

Abstract: Power Quality has become one of the main measures of distribution network operation evaluation. Utilities and customers monitor the different PQ aspects for benchmarking and compatibility check. This is a second part of a paper that investigates a long term PQ measurement campaign. In this part, global PQ indices are applied to combine a number of PQ phenomena indices into one number that represents the PQ performance of a site. Two global PQ indices are applied on 64-week PQ measurements of 8 sites. The indices are compared in terms of the final rank of sites. Different levels of flexibility are introduced to the indices to consider a variable importance or priority of the phenomena and sites.

Frequency Stabilisation Using VSC-HVDC

Abstract: Frequency control of AC systems using Voltage Source Converter (VSC) HVDC links is of growing importance as interconnectors and links to offshore generation replace conventional synchronous generators in many countries. This paper investigates the use of local system feedback and advanced droop control to achieve frequency stabilization for a two-area test system.

Frequency stabilisation using VSC-HVDC

Abstract: Frequency control of AC systems using Voltage Source Converter (VSC) HVDC links is of growing importance as interconnectors and links to offshore generation replace conventional synchronous generators in many countries. This paper investigates the use of local system feedback and advanced droop control to achieve frequency stabilization for a two-area test system.

Probabilistic ranking of power system loads for voltage stability studies in networks with renewable generation

Abstract: Power system voltage stability is a crucial aspect of power system study, and loads have a significant influence on it. This paper uses PV curves to rank the power system loads according to their influence on the voltage stability. The Monte Carlo simulation is used to generate uncertainties to reflect the stochastic behaviour of power systems. The ranking order is acquired after considering different loading conditions obtained from the annual loading curve. The distribution of critical points on the ‘PV nose curve’ for the most important loads and least important loads in the network are used to verify the ranking. Different load models are applied to investigate the effect of load models on the critical point positions. The results show that load models have significant influence on the voltage collapse point.

Feasibility study of applicability of recurrence quantification analysis for clustering of power system dynamic responses

Abstract: A methodology based on Recurrence Quantification Analysis (RQA) for the clustering of generator dynamic behavior is presented. RQA is a nonlinear data analysis method, which is used in this paper to extract features from measured generator rotor angle responses that can be used to cluster generators in groups with similar oscillatory behavior. The possibility of extracting features relevant to damping and frequency of oscillations present in power systems is studied. The k-Means clustering algorithm is further used to cluster the generator responses in groups exhibiting well or poorly damped oscillations, based on the extracted features from RQA. The effectiveness of RQA is investigated using simulated responses from a modified version of the IEEE 68 bus network, including renewable energy resources.

Application of sequential testing problem to online detection of transient stability status for power systems

Abstract: We address the problem of predicting the transient stability status of a power system as quickly as possible in real time subject to probabilistic risk constraints. The goal is to minimise the average time taken after a fault to make the prediction, and the method is based on ideas from statistical sequential analysis. The proposed approach combines probabilistic neural networks with dynamic programming. Simulation results show an approximately three-fold increase in prediction speed when compared to the use of pre-committed (fixed) prediction times.

Towards application of text mining for enhanced power network data analytics - Part I: Retrieval and ranking of textual data from the internet

Abstract: This paper presents initial results of methodologies developed for extracting useful information from online textual data in the field of power networks. Since there are no tools developed specifically for power engineering, an attempt was made to take advantage of the existing tools for specialized web browsing and data extraction. Two methodologies are explored: one for extracting on-line documents (journal papers, technical reports, etc.) that are highly related to a specific topic, and the second one for extracting highly related sentences, providing a literature summary on the topic. The first results are promising, showing great potential for text-mining applications as a part of power networks data analytics.

Optimal compensation of transmission lines based on minimisation of the risk of subsynchronous resonance

Abstract: The paper proposes a framework for the selection of optimal configuration of line compensation devices, thyristor controlled series capacitors (TCSC) and conventional fixed series capacitors, for reducing the risk of subsynchronous resonance (SSR) in the network while maintaining required power transfer. The methodology developed in the paper is based on the robust risk evaluation of SSR that takes into consideration the severity of subsynchronous resonance and probability of its occurrence. The subsynchronous resonance risk index developed previously is used to assess the severity of subsynchronous resonance. The line outage model is employed to determine the probability of contingencies potentially leading to subsynchronous resonance. It is demonstrated that with relatively small participation of TCSC, even with the most basic control, in series compensation of lines the risk of SSR can be successfully managed

Methodology for close to real time profiling of aggregated demand using data streams from smart meters

Abstract: This paper discusses potential improvement in accuracy of estimation of load profiles at substation/aggregation point if the demand data is collected directly from smart meters rather than from balancing meters at bulk supply points. It proposes a bottom-up approach for development of daily load curves for domestic load sector by aggregating data coming as real-time data series from smart meters. In order to illustrate the concepts an assumption is made that all the smart meters in an area have the ability to measure instantaneous real power demand of each individual appliance. Following this, a probabilistic bottom-up approach is applied to generate reactive power demand at the point of aggregation. It is further assumed that the collected data streams have different sampling steps and that there are some missing data in recorded data streams. Different data conditioning methods are used to investigate the accuracy of demand aggregation at different aggregation levels not only in terms of total demand but also in terms of demand categories and controllable and uncontrollable demand.

Application of data analytics for information retrieval from a typical DSO's database

Abstract: This paper introduces the reasons for big data analytics in distribution network studies and potential benefits it could give. Summary of the most common data mining methods used in power system studies is also given, followed by a comparative analysis. A use case is shown at the end in order to present some examples of extraction of useful information from raw data stored in a real distribution utility's database. This was done by using some of the basic data mining methods applied to different types of attributes describing distribution system feeders in 11 kV and 6.6 kV network. The initial results showed that the usefulness of information depends on the level of data aggregation, as well as the choice of data analytics method.

Probabilistic ranking of critical parameters affecting voltage stability in network with renewable generation

Abstract: This paper introduces a probabilistic method for the ranking of influential uncertain parameters for the accurate assessment of power system voltage stability. Future power systems will be highly interconnected and complex with a variety of uncertain parameters such as the injection of intermittent renewable energy resources, the adoption of flexible hierarchical control structures and the appearance of new types of loads. Identifying and ranking the uncertain parameters are important in future power system operations since they can provide referable indexes for system operators to achieve better system management with less monitoring. This paper presents the probabilistic method for the identification and ranking of critical uncertain parameters. A modified version of the 68 bus NETS-NYPS test system is used in this study for simulation studies. The effects of uncertain parameters are modelled with Monte-Carlo method in the environments of MATLAB and DIgSILENT PowerFactory. The performances of the ‘nose-point area’ of P-V Curves for system load buses are used as indexes when evaluating their sensibility for specific uncertainties.

Comparative analysis of data availability and data requirements for efficient management and control of future distribution networks

Abstract: This paper starts by discussing the data requirements for efficient control and management of future power networks and justifies the need for different types of data. It then compares the data requirements with presently available or routinely collected data by distribution system operators (DSOs) and identifies the mismatch between the two. Finally, considering growing deployment and increasing reliance on smart meters, it presents an overview of the technical specifications of typical commercially available smart meters and discusses the possibilities of information retrieval from the data provided by smart meters. The results of this analysis illustrate to what extent the existing smart meters can support demand side management (DSM) and identify what additional functionalities smart meters should have in order to facilitate envisaged services.

Towards application of text mining for enhanced power network data analytics - Part II: Offline analysis of textual data

Abstract: Text mining is a subdivision of data mining technologies used to extract useful information from unstructured textual data. In recent years, power distribution networks have become more complex due to the versatile consumer demand and integration of distributed energy resources. This has led to the need for enhanced data processing and analysis, i.e., data analytics, in distribution system studies. This paper for the first time explores the feasibility of application of text mining methods as a part of power system data analytics. The focus is on identifying and describing the steps that need to be taken for the knowledge extraction from large offline textual document collections and on demonstrating the effectiveness of the whole process if undertaken by a power system engineer, i.e., a nonspecialist in the area of text mining.