Sasa Djokic - Academia.edu (original) (raw)

Papers by Sasa Djokic

IEEE Transactions on Power Delivery, 2006

Control and limitation of harmonic current emission from new customer installations is essential ... more Control and limitation of harmonic current emission from new customer installations is essential for maintaining supply voltage distortions below the prescribed compatibility or planning levels, and is also required for ensuring electromagnetic compatibility. At the planning stage, emissions of low-order harmonic current are evaluated based on several assumptions and simplifications related to, e.g. use of summation exponent and presence of background distortion. Consequently, the actual contribution of a new customer installation to the resulting supply voltage distortion after its connection can significantly differ from the estimation at the planning stage, what typically results in a less-efficient utilization of the available grid hosting capacity. A new CIGRE/CIRED Joint Working Group C4.42 is established to identify or develop suitable measurement-based methods for reliable and accurate assessment of the harmonic contributions from new customer installations to the resulting supply voltage distortion. One of the initial tasks of JWG C4.42 is comparison of existing methods, which is presented in this paper.

International Power Electronics and Motion Control Conference, Sep 6, 2010

This paper proposes a methodology for improving the representation of component-based models inte... more This paper proposes a methodology for improving the representation of component-based models intended for harmonic analysis of generic aggregate LED lamp loads by deriving pdfs of the parameter values. The methodology is illustrated using time-domain current waveform measurements of LED lamps with switch-mode driver circuits (without power factor correction). Following the model derivation, Monte Carlo simulations are used to synthesize generic aggregate LED lamp loads, which are compared against random aggregations of 51 measured lamps. It is demonstrated that a single model with mean values of the parameters, scaled by the total power demand, can be used to simplify the representation of generic aggregate LED lamp loads in large-scale emission assessment studies. Correction factors, which reduce the effect of this simplification in frequency domain applications, are also derived and discussed.

The management of static security constraints (bus voltages and branch thermal limits) is one of ... more The management of static security constraints (bus voltages and branch thermal limits) is one of the continuous tasks performed by operators. These constraints at any operating point can be divided into noncritical and critical constraints. To be secure, the system must fulfil both types of security constraints at all the time. While primary active and reactive power controls can manage noncritical constraints, emergency controls or remedial actions must be computed using optimal power flow (OPF) analysis to alleviate critical constraint violations. If the post contingency network involves a violation of critical constraints (which are not known a priori), conventional OPF algorithms may fail to converge and compute required control action, as the corresponding optimization problem becomes mathematically infeasible. Hence, the identification of critical constraint violations is necessary, as their location in the network decides the installation (during planning) and activation (during operation) of emergency controls. In this context, this paper first identifies the location of critical constraint violations using a metaheuristic approach, and then proposes a method to optimally place or control distributed generation to alleviate network thermal and voltage congestions during emergency situations.

This paper analyses the equipment susceptibility to multiple voltage dips and short interruptions... more This paper analyses the equipment susceptibility to multiple voltage dips and short interruptions, which often originate from switching time sequences of automation and reconfiguration actions adopted in smart grids. The paper introduces a general trellis model, which allows to calculate the relevant quantities of equipment response in terms of shutdown probability to the sequences of voltage dips of assigned characteristics, assuming that the hypotheses of equipment “lack of memory” are true. The presence of uncertain shutdown regions in dip-tolerance diagrams based on residual voltage and duration of individual voltage dips is taken into account. Numerical analyses demonstrate the applicability and the usefulness of the proposed models to evaluate customer reliability and power quality performance.

This paper presents the results from the work of WG C4.110, a joint working group by CIGRE, CIRED... more This paper presents the results from the work of WG C4.110, a joint working group by CIGRE, CIRED and UIE. Its mandate period stretched from early 2006 through early 2009. The group has addressed several aspects of the immunity of, especially, industrial equipment against voltage dips. Compared to the work earlier groups, the equipment performance is not seen as a final aim, but as a step towards the ultimate aim: allowing the process to ride through the voltage dip. Some of the contributions and conclusions from the WG C4.110 are discussed in this paper, ✓ Description of voltage dips, ✓ Equipment and process immunity, ✓ Testing and characterization, ✓ Economics, ✓ Immunity classes and application, ✓ Further work.

This paper investigates significant changes in the performance of switch-mode power supplies used... more This paper investigates significant changes in the performance of switch-mode power supplies used in standard desktop PCs (PC-SMPS'), which might occur when a PC-SMPS transfers from high to low operating powers. Using the experimental results, it is shown that due to a partial discharging of dc link capacitor at low powers, the input ac current of tested PC-SMPS' loses its periodicity with supply voltage frequency (i.e. line frequency). In such cases, there is significant increase of PC-SMPS' current waveform distortion and substantial decrease of efficiency and operational power factors. In existing literature, this phenomena is analysed in terms of electrical and control circuit “instability”, denoted as a slow-scale bifurcation and chaotic operation, where the main concern was circuit design (or re-design) for achieving full, or at least improved operational stability. In this paper, however, lost periodicity is analysed from both metrological and analytical points of view, focusing on appropriate measurement and calculation procedures for evaluating changes in PC-SMPS' performance (efficiencies and power factors) and power quality characteristics (waveform distortion). The presented analysis can be applied for both testing and evaluating performance of larger SMPS' and other power electronic devices with a similar behaviour (e.g. PV inverters).

In this paper, a component-based modelling methodology is applied to the commercial load sector. ... more In this paper, a component-based modelling methodology is applied to the commercial load sector. This approach allows for greater insight into the temporal variations in the electrical characteristics of the commercial load sector than currently available modelling practices. This is illustrated by analyzing the half hourly changes in the voltage dependency of the demand and harmonic distortion of a generic UK commercial load sector representation; however, the methodology can be applied wherever similar datasets exist. The results highlight differences between on-peak and off-periods, and provide a useful starting place for probabilistic model development.

Arabian journal for science and engineering, Mar 31, 2020

Load rebound characteristics (LRC), which means that restoration of distribution systems with lar... more Load rebound characteristics (LRC), which means that restoration of distribution systems with large thermostatically controlled load (TCL) after an outage may create significantly higher load demand than normal (namely the demand if this outage doesn't occur), will make the reliability index-expected energy not supplied (EENS) of distribution systems obtained by traditional methods pessimistic and inaccurate. Since EENS is one of the base points of investment planning for distribution system operator, it is necessary to consider the LRC in the reliability evaluation and provide a more accurate evaluation on EENS. This paper proposed a new reliability evaluation method of smart distribution systems based on a non-negative k-singular value decomposition (NN-K-SVD) algorithm and a sequential Monte Carlo (SMC) simulation. The NN-K-SVD algorithm is used to identify the TCL in total load, and a reliability evaluation model considering the LRC is established based on an electric water heater model. The SMC simulation is used to mimic the operation of distribution systems based on the established reliability evaluation model, and calculate reliability indices. Furthermore, the feasibility of the proposed method is validated by extensive cases studies.

Assessing distortion within the lEC framework in the presence of high frequency (HF) components i... more Assessing distortion within the lEC framework in the presence of high frequency (HF) components is considered in this paper from the perspective of signal processing. To reduce uncertainty when assessing HF distortion, the lEC suggests the use of a high-pass filter, which can be analogue (with a dedicated channel for HF assessment) or integrated in its digital form into the signal processing stage. This paper extends the proven desynchronized processing technique (DPT), previously introduced for harmonic and interharmonic analysis, as an alternative filtering technique and compares its performance against three digital filtering techniques presented in literature. Particular attention is paid to the error introduced by the desynchronization of the fundamental frequency, which naturally deviates from its rated value, with the fixed 200ms window recommended by the lEC for HF distortion assessment. Two case studies highlight the strengths and weaknesses of the techniques and demonstrate the potential of the DPT to analyse HF distortion.

This paper presents a non-homogeneous Markov Chain (MC) model for generation of wind speed (WS) a... more This paper presents a non-homogeneous Markov Chain (MC) model for generation of wind speed (WS) and wind direction (WD) synthetic time series taking into account their daily, monthly and seasonal characteristics. The bivariate nature of the wind process, represented by WS and WD, is modelled by means of an equivalent univariate random variable W, capable of taking into account the statistical dependency existing between WS and WD. A statistical characterization of the wind energy resource at the specific considered site demonstrates the time non-stationarity of the wind process over the year and over the seasons, so twelve monthly transition probability matrices of the variable W are developed. One thousand synthetic time series, each of three years length, are generated in a Monte Carlo framework, demonstrating the excellent performances and overall robustness of the presented model, also using new non-conventional metrics based on Markov transition matrices.

This paper evaluates deterministic and probabilistic approaches for assessing hosting capacity (H... more This paper evaluates deterministic and probabilistic approaches for assessing hosting capacity (HC) of distribution networks for wind-based distributed generation (DG). The presented methodology considers variations of demands and DG power outputs, as well as dynamic thermal ratings (DTR) of network components. Deterministic approaches are based on a limited number of scenarios with minimum and maximum demands and DTR limits, while probabilistic approaches use simultaneous hourly values of all input parameters. The presented methodology has three stages. First, locational HC (LHC) of individual buses is calculated assuming connection of a single DG unit in the considered network. Afterwards, the LHC results are used to calculate network HC (NHC), assuming that DG units are connected at all network buses. Finally, busto-bus LHC-sensitivity factors are used to determine LHC and NHC for any number of DG units connected at arbitrary network buses.

As a significant increase of electric vehicles (EV) is expected in the future, studies of the imp... more As a significant increase of electric vehicles (EV) is expected in the future, studies of the impact of this new technology on the harmonic levels in distribution grids are very important. Realistic simulation results require reliable and accurate models of EV chargers, which generally utilize complex circuits including control algorithms for the active PFC. As part of a research project on the impact of high EV penetrations on the harmonic levels in public low voltage networks, a time-domain and a frequency-domain approach for the design of such models have been studied and compared in terms of difficulty, development effort and accuracy. The paper discusses in detail the methodology for the development of both types of models and summarizes their specific advantages and drawbacks, for both assessing performance of individual devices and for analyzing their interactions in network simulations. Based on different objectives, the conclusions provide guidance on the preferred modelling approaches for efficient and reliable network simulations.

IEEE Transactions on Power Delivery, 2006

Control and limitation of harmonic current emission from new customer installations is essential ... more Control and limitation of harmonic current emission from new customer installations is essential for maintaining supply voltage distortions below the prescribed compatibility or planning levels, and is also required for ensuring electromagnetic compatibility. At the planning stage, emissions of low-order harmonic current are evaluated based on several assumptions and simplifications related to, e.g. use of summation exponent and presence of background distortion. Consequently, the actual contribution of a new customer installation to the resulting supply voltage distortion after its connection can significantly differ from the estimation at the planning stage, what typically results in a less-efficient utilization of the available grid hosting capacity. A new CIGRE/CIRED Joint Working Group C4.42 is established to identify or develop suitable measurement-based methods for reliable and accurate assessment of the harmonic contributions from new customer installations to the resulting supply voltage distortion. One of the initial tasks of JWG C4.42 is comparison of existing methods, which is presented in this paper.

International Power Electronics and Motion Control Conference, Sep 6, 2010

This paper proposes a methodology for improving the representation of component-based models inte... more This paper proposes a methodology for improving the representation of component-based models intended for harmonic analysis of generic aggregate LED lamp loads by deriving pdfs of the parameter values. The methodology is illustrated using time-domain current waveform measurements of LED lamps with switch-mode driver circuits (without power factor correction). Following the model derivation, Monte Carlo simulations are used to synthesize generic aggregate LED lamp loads, which are compared against random aggregations of 51 measured lamps. It is demonstrated that a single model with mean values of the parameters, scaled by the total power demand, can be used to simplify the representation of generic aggregate LED lamp loads in large-scale emission assessment studies. Correction factors, which reduce the effect of this simplification in frequency domain applications, are also derived and discussed.

The management of static security constraints (bus voltages and branch thermal limits) is one of ... more The management of static security constraints (bus voltages and branch thermal limits) is one of the continuous tasks performed by operators. These constraints at any operating point can be divided into noncritical and critical constraints. To be secure, the system must fulfil both types of security constraints at all the time. While primary active and reactive power controls can manage noncritical constraints, emergency controls or remedial actions must be computed using optimal power flow (OPF) analysis to alleviate critical constraint violations. If the post contingency network involves a violation of critical constraints (which are not known a priori), conventional OPF algorithms may fail to converge and compute required control action, as the corresponding optimization problem becomes mathematically infeasible. Hence, the identification of critical constraint violations is necessary, as their location in the network decides the installation (during planning) and activation (during operation) of emergency controls. In this context, this paper first identifies the location of critical constraint violations using a metaheuristic approach, and then proposes a method to optimally place or control distributed generation to alleviate network thermal and voltage congestions during emergency situations.

This paper analyses the equipment susceptibility to multiple voltage dips and short interruptions... more This paper analyses the equipment susceptibility to multiple voltage dips and short interruptions, which often originate from switching time sequences of automation and reconfiguration actions adopted in smart grids. The paper introduces a general trellis model, which allows to calculate the relevant quantities of equipment response in terms of shutdown probability to the sequences of voltage dips of assigned characteristics, assuming that the hypotheses of equipment “lack of memory” are true. The presence of uncertain shutdown regions in dip-tolerance diagrams based on residual voltage and duration of individual voltage dips is taken into account. Numerical analyses demonstrate the applicability and the usefulness of the proposed models to evaluate customer reliability and power quality performance.

This paper presents the results from the work of WG C4.110, a joint working group by CIGRE, CIRED... more This paper presents the results from the work of WG C4.110, a joint working group by CIGRE, CIRED and UIE. Its mandate period stretched from early 2006 through early 2009. The group has addressed several aspects of the immunity of, especially, industrial equipment against voltage dips. Compared to the work earlier groups, the equipment performance is not seen as a final aim, but as a step towards the ultimate aim: allowing the process to ride through the voltage dip. Some of the contributions and conclusions from the WG C4.110 are discussed in this paper, ✓ Description of voltage dips, ✓ Equipment and process immunity, ✓ Testing and characterization, ✓ Economics, ✓ Immunity classes and application, ✓ Further work.

This paper investigates significant changes in the performance of switch-mode power supplies used... more This paper investigates significant changes in the performance of switch-mode power supplies used in standard desktop PCs (PC-SMPS'), which might occur when a PC-SMPS transfers from high to low operating powers. Using the experimental results, it is shown that due to a partial discharging of dc link capacitor at low powers, the input ac current of tested PC-SMPS' loses its periodicity with supply voltage frequency (i.e. line frequency). In such cases, there is significant increase of PC-SMPS' current waveform distortion and substantial decrease of efficiency and operational power factors. In existing literature, this phenomena is analysed in terms of electrical and control circuit “instability”, denoted as a slow-scale bifurcation and chaotic operation, where the main concern was circuit design (or re-design) for achieving full, or at least improved operational stability. In this paper, however, lost periodicity is analysed from both metrological and analytical points of view, focusing on appropriate measurement and calculation procedures for evaluating changes in PC-SMPS' performance (efficiencies and power factors) and power quality characteristics (waveform distortion). The presented analysis can be applied for both testing and evaluating performance of larger SMPS' and other power electronic devices with a similar behaviour (e.g. PV inverters).

In this paper, a component-based modelling methodology is applied to the commercial load sector. ... more In this paper, a component-based modelling methodology is applied to the commercial load sector. This approach allows for greater insight into the temporal variations in the electrical characteristics of the commercial load sector than currently available modelling practices. This is illustrated by analyzing the half hourly changes in the voltage dependency of the demand and harmonic distortion of a generic UK commercial load sector representation; however, the methodology can be applied wherever similar datasets exist. The results highlight differences between on-peak and off-periods, and provide a useful starting place for probabilistic model development.

Arabian journal for science and engineering, Mar 31, 2020

Load rebound characteristics (LRC), which means that restoration of distribution systems with lar... more Load rebound characteristics (LRC), which means that restoration of distribution systems with large thermostatically controlled load (TCL) after an outage may create significantly higher load demand than normal (namely the demand if this outage doesn't occur), will make the reliability index-expected energy not supplied (EENS) of distribution systems obtained by traditional methods pessimistic and inaccurate. Since EENS is one of the base points of investment planning for distribution system operator, it is necessary to consider the LRC in the reliability evaluation and provide a more accurate evaluation on EENS. This paper proposed a new reliability evaluation method of smart distribution systems based on a non-negative k-singular value decomposition (NN-K-SVD) algorithm and a sequential Monte Carlo (SMC) simulation. The NN-K-SVD algorithm is used to identify the TCL in total load, and a reliability evaluation model considering the LRC is established based on an electric water heater model. The SMC simulation is used to mimic the operation of distribution systems based on the established reliability evaluation model, and calculate reliability indices. Furthermore, the feasibility of the proposed method is validated by extensive cases studies.

Assessing distortion within the lEC framework in the presence of high frequency (HF) components i... more Assessing distortion within the lEC framework in the presence of high frequency (HF) components is considered in this paper from the perspective of signal processing. To reduce uncertainty when assessing HF distortion, the lEC suggests the use of a high-pass filter, which can be analogue (with a dedicated channel for HF assessment) or integrated in its digital form into the signal processing stage. This paper extends the proven desynchronized processing technique (DPT), previously introduced for harmonic and interharmonic analysis, as an alternative filtering technique and compares its performance against three digital filtering techniques presented in literature. Particular attention is paid to the error introduced by the desynchronization of the fundamental frequency, which naturally deviates from its rated value, with the fixed 200ms window recommended by the lEC for HF distortion assessment. Two case studies highlight the strengths and weaknesses of the techniques and demonstrate the potential of the DPT to analyse HF distortion.

This paper presents a non-homogeneous Markov Chain (MC) model for generation of wind speed (WS) a... more This paper presents a non-homogeneous Markov Chain (MC) model for generation of wind speed (WS) and wind direction (WD) synthetic time series taking into account their daily, monthly and seasonal characteristics. The bivariate nature of the wind process, represented by WS and WD, is modelled by means of an equivalent univariate random variable W, capable of taking into account the statistical dependency existing between WS and WD. A statistical characterization of the wind energy resource at the specific considered site demonstrates the time non-stationarity of the wind process over the year and over the seasons, so twelve monthly transition probability matrices of the variable W are developed. One thousand synthetic time series, each of three years length, are generated in a Monte Carlo framework, demonstrating the excellent performances and overall robustness of the presented model, also using new non-conventional metrics based on Markov transition matrices.

This paper evaluates deterministic and probabilistic approaches for assessing hosting capacity (H... more This paper evaluates deterministic and probabilistic approaches for assessing hosting capacity (HC) of distribution networks for wind-based distributed generation (DG). The presented methodology considers variations of demands and DG power outputs, as well as dynamic thermal ratings (DTR) of network components. Deterministic approaches are based on a limited number of scenarios with minimum and maximum demands and DTR limits, while probabilistic approaches use simultaneous hourly values of all input parameters. The presented methodology has three stages. First, locational HC (LHC) of individual buses is calculated assuming connection of a single DG unit in the considered network. Afterwards, the LHC results are used to calculate network HC (NHC), assuming that DG units are connected at all network buses. Finally, busto-bus LHC-sensitivity factors are used to determine LHC and NHC for any number of DG units connected at arbitrary network buses.

As a significant increase of electric vehicles (EV) is expected in the future, studies of the imp... more As a significant increase of electric vehicles (EV) is expected in the future, studies of the impact of this new technology on the harmonic levels in distribution grids are very important. Realistic simulation results require reliable and accurate models of EV chargers, which generally utilize complex circuits including control algorithms for the active PFC. As part of a research project on the impact of high EV penetrations on the harmonic levels in public low voltage networks, a time-domain and a frequency-domain approach for the design of such models have been studied and compared in terms of difficulty, development effort and accuracy. The paper discusses in detail the methodology for the development of both types of models and summarizes their specific advantages and drawbacks, for both assessing performance of individual devices and for analyzing their interactions in network simulations. Based on different objectives, the conclusions provide guidance on the preferred modelling approaches for efficient and reliable network simulations.