Y Yu Xiang

Optimization of State-Estimator-Based Operation Framework Including Measurement Placement for Medium Voltage Distribution Grid

To provide a guideline for grid operators to manage medium voltage distribution grids, this paper presents a probabilistic approach for measurement placement optimization and a state-estimator-based operation framework. The Monte Carlo method is used to optimize the measurement placement with minimum measured points, which support the observability of state estimator and satisfy the requirements for operation activities. The proposed framework covers several important factors from practical situations, including cost allocation of measurement systems, unavailability of measurement data, and small voltage deviation across medium voltage (MV) grids. Finally, a case study on a typical European distribution grid demonstrates the feasibility of the framework.

Statistical error analysis of household load profile for medium voltage grid state estimation

The paper comprises a detailed analysis of the error ranges of two commonly proposed load estimation methods for household loads, based on statistics. Firstly, a household load profile model is built taking individual user preference into account. To analyze the error range, it is divided into system error (representing proportion factors) and random error (representing the diversity of stochastic loads and measurement inaccuracies). For both estimation methods, two kinds of bounds are used to evaluate the error range, with central limit theorem. Finally, comparisons between different errors are made using field data from a Dutch utility.

Impact of residual harmonic current on operation of residual current devices

Residual current devices (RCDs) are most commonly used as a part of the protection system in low voltage installations. RCDs are supposed to disconnect the supply whenever the imbalance current between the live conductor(s) and the neutral conductor, i.e. residual current, exceeds the predetermined value. Such an imbalance is sometimes caused by current leakage through the body of a person who is grounded and accidentally touching the energized part of the circuit, which could result in a lethal shock. However, harmonic components in ground fault current may influence the tripping sensitivity of RCDs, and the protection against electrical shock may not be effective. In the paper, results of theoretical analysis and experiments are presented. Several mechanisms which could influence the tripping sensitivity of RCDs are validated, and their dominant conditions are identified.

Voltage dip state estimation in distribution networks by applying Bayesian inference

The performance of the system regarding voltage dips is commonly characterized with the SARFIx index, which gives the average frequency of voltage dips. Due to the limited number of measurement points, the actual residual voltages of many nodes are not taken into account when the index is calculated. As its impossible to measure the dip level at every node of a feeder, the voltage dips occurring at nonmonitored nodes should be estimated from the values recorded at monitored nodes with the consideration of measurement accuracy. This paper proposes a voltage dip state estimation method based on Bayesian inference. The performance of the proposed method is assessed through case study applied in a typical distribution network. Monte Carlo simulation is used to obtain the statistic results. The proposed method shows the flexibility of different measurement quantities and the adequacy for the analysis of distribution networks.

Assessment of the voltage level and losses with photovoltaic and electric vehicle in low voltage network

Livelab from Alliander, a network operator, is a program which started to measure electrical and power quality data in the Dutch distribution network since 2013. A proper probability distribution can be used to model load distribution on feeders. This paper presents a methodology to generate the residential load profiles by cumulative distribution function based on Livelab measured data. Applying Monte Carlo simulation, a probabilistic three phase Load Flow program is developed to investigate the voltage level and network losses (where harmonic loss is considered) of a typical low voltage network with the consideration of variable loads and random locations. The prospective impact to voltage level and losses due to the grid- connected photovoltaic power generation systems and electrical vehicles are discussed. Monte Carlo simulations are carried out with the different penetration rates of photovoltaic generation and electrical vehicles, maximum capacity of grid-connected equipment can be estimated and the possible network losses are calculated.

Probabilistic analysis on measurement timescale in medium voltage grid

The paper presents a probabilistic approach to evaluate the errors due to measurement timescale in medium voltage distribution grid. Three common data acquisition methods by measurement devices are shown, and two kinds of measurement errors due to timescale are defined. Based on Gaussian Process, the stochastic characteristics of household loads are discussed, which are important random variables influencing the grid. A probabilistic approach with Monte Carlo method is presented to evaluate the measurement errors due to timescale problem. Finally, a case study on a simplified grid is given.

Bayesian-Inference-Based Voltage Dip State Estimation

Voltage dip state estimation (VDSE) tries to estimate the voltage dip characteristics at nonmonitored buses from measured voltage dip values at monitored buses. In this paper, the VDSE is addressed through the method based on Bayesian inference. A priori including the fault position among other grid conditions is used to estimate the residual voltage at each bus based on the measurement quantities, including their uncertainties. The dip duration is calculated with the time setting of protection system incorporating the uncertainties due to dip detection algorithm of the root mean square values. The proposed method has been applied to the IEEE 13-bus and IEEE 123-bus distribution test systems for multiple simulation scenarios, such as with or without distributed generation and different types of faults. The simulation results show good observability of the network.

Influence of high frequency current harmonics on (Smart) energy meters

(Smart) meters which use Rogowski coils are verified to be sensitive for the high frequency signals. This will lead to incorrect records of consumers energy consumption. Power electronic devices are the possible harmonic sources. In this paper, the output currents of photovoltaic (PV) systems are measured in the range 2-150kHz and the high frequency harmonic currents are regenerated in the laboratory to investigate their influence on the smart energy meters. It is realized that the insufficient dynamic range in the meters could be the possible reason of inaccuracy. After the experiments, it is concluded that the high frequency current can make the output signal of the Rogowski coil beyond the dynamic range of smart meters and sequentially less registered energy.

Improving operation of medium voltage grid with state estimator and advanced metering system

This paper adopts a state estimator and a meter placement approach to improve the operation of medium voltage distribution grids. In the framework different input specifications are designated. Moreover, three common activities during normal operation are analyzed in detail: voltage regulation, switching consequence assessment, and loss estimation. To evaluate the performance of these activities, several indices are defined. To demonstrate the improvements that the framework may gain over the currently used operation strategy, an extensive case study is performed on a typical European distribution grid.

Voltage sag estimation based on probabilistic short circuit including the fault ride through effect of distributed generator

Short circuit (SC) fault is one of the most severe disturbances faced by the power system. SC faults not only lead to high SC currents but also cause voltage sag in a wide area. In this paper the voltage sag problem caused by SC fault are estimated based on the probabilistic SC concept. The research is based on a typical Dutch Medium Voltage distribution grid where the penetration of Distributed Generation (DG) is experiencing rapid growth. Fault Ride Through (FRT) capability will be obligatory for all types of DG system in future grid code which is different with currently situation. However, SC contribution from Voltage Source Converter (VSC) based DG is neglected according to IEC 60909 standard and there is no available model to consider FRT behavior of VSC-DG yet in the simulation software DIgSILENT PowerFactory. Therefore an equivalent model is developed in PowerFactory in order to emulate VSC-DGs FRT capability and study its effect on the voltage sag estimation results. The simulation results show that the probability of occurrence of severe voltage sag is reduced and the system overall voltage sag performance is improved both due to the voltage support effect during FRT.