In recent years, harmonic current levels have increased in residential distribution networks due to the growing penetration of nonlinear loads at customer premises. To properly model the harmonic current injection of residences, probabilistic models that account for the stochastic behavior of domestic loads are needed. As a part of the study reported in this paper, a field measurement campaign was performed in 24 dwellings during a one-week period with a one-minute resolution. Harmonic current magnitudes and phase angles of odd harmonic orders up to the 25th as well as the active and reactive power demanded were recorded. By applying two different unsupervised learning techniques (i.e., nonparametric density estimation and Gaussian mixture models) to these measurements, two probabilistic models of harmonic injection of individual residential sites have been derived. Statistical probability distributions have been determined for the magnitude and phase of each harmonic order segmented in different power demand intervals. The developed models are validated on a test feeder by comparing harmonic voltages caused by the injection of the synthetically generated currents with those obtained from measurements. Both field measurement results and the detailed data defining the probabilistic harmonic current models are made public in an Open Science database repository.

The generalized adoption of power-electronic-based loads in residential networks is leading to increasing concern about harmonic distortion levels. At the same time, residential distribution systems are insufficiently monitored to provide a reliable analysis of harmonic limits compliance at the point of common coupling of customers. To address this problem, this paper proposes a practical method to estimate voltage harmonic levels in real time at all connection points of residential consumers. The proposed methodology assumes unknown network parameters and topology according to the limited data availability which is usual in low voltage residential networks. The proposed harmonic estimation method uses information gathered from conventional smart meters typically installed in the network (which register fundamental voltage and power) and from a reduced number of installed power quality monitors. A penetration of power quality meters below 10% has proved to provide accurate estimations of voltage harmonic levels at the whole network. The paper also proposes a methodology to optimally locate these power quality meters in the network. Validation of the methodology is performed in two different European distribution test networks with 55 buses and 471 obtaining satisfactory results. The method’s robustness and the parameters affecting accuracy are also analyzed in the article.

Voltage sags are power quality disturbances mainly caused by faults with a highly random nature. For this reason, the severity and frequency of occurrence of voltage sags is typically analyzed by means of probabilistic methods, such as Monte Carlo simulations, where different probability distributions are selected to model the uncertain inputs (e.g., fault impedance, fault location, fault rate, etc.). The selection and parameterization of the appropriate probability functions involves many difficulties to ensure that they represent the distribution of input variables in a realistic way and so, many different approaches can be found in literature. This paper studies the sensitivity of voltage sags severity to the uncertainty in the input models in order to focus modeling effort on the input variables that are more critical and effectively have an influence on voltage sags levels, whereas non-influential parameters can be neglected or modeled by means of simplified hypothesis. The method proposed is based on Design of Experiments and Analysis of Variance (ANOVA) and is able to discriminate with a statistic confidence level whether the uncertainty in the input significantly affects voltage sags severity indices or not. The effect of uncertainty on voltage sags indices SARFI90 and SARFI70 has been evaluated in realistic scenarios for IEEE test networks of 24 and 118 buses and in the power system of Ecuador with 357 nodes. General trends have been also established that help to understand the effect of the modeling of input parameters on the estimated number of sags.

The accelerating transition toward sustainable transportation is expected to significantly increase the penetration of electric vehicles (EVs) into distribution networks. EVs, as nonlinear loads with a AC-to-DC convertors, introduce harmonic distortion that may exceed regulatory limits or impact power system components. Accurate prediction of resulting voltage harmonics requires probabilistic models capturing the non-linearity and inherent variability of EV harmonic emissions. In this work, an extensive field measurement campaign was conducted to record harmonic current phasors for odd harmonic orders up to the 50th, across more than 54 charging cycles from 14 EVs representing nine different models. The measurements show significant differences in harmonic injection profiles across EVs, attributed to variations in design, power electronic, and charging technologies. A Multivariate Gaussian Mixture Model (GMM) is applied to characterize EV harmonic current injections and to generate synthetic samples consistent with observed distributions. The proposed model accurately captures the underlying patterns and correlations between harmonic magnitudes and angles. Model validation has been performed by comparing the probability distributions obtained from measured and synthetically generated harmonic currents, as well as the harmonic voltage levels resulting in a distribution feeder. The field measurement data has been made publicly available to support further research.

This paper proposes a methodology to assess the influence of transformer rated power on the power quality indices at low voltage residential networks. Taking into account the stochastic nature of residential loads and the possibility of photovoltaic generation, a Monte-Carlo simulation is performed in order to account for the probabilistic aspects of the problem.In this work, different transformer ratings are simulated and the effect on power quality parameters, such as voltage unbalance and voltage total harmonic distortion (THD), is calculated under probabilistic basis. These power quality indices are compared to the compliance limits proposed in standard EN50160.The work proposed in this paper allows predicting power quality affection caused by changes in the ratio between transformer rated power and network aggregated demand, as well as establishing general conclusions regarding desirable transformer size in order to not to exceed power quality limits.