Publications

Topology identification in transmission systems has historically been accomplished using SCADA measurements. In distribution systems, however, SCADA measurements are insufficient to determine system topology. An accurate system topology is essential for distribution system monitoring and operation. Recently there has been a proliferation of Advanced Metering Infrastructure (AMI) by the electrical utilities, which improved the visibility into distribution systems. These measurements offer a unique capability for Distribution System Topology Identification (DSTI). A novel approach to DSTI is presented in this paper which utilizes the voltage magnitudes collected by distribution grid sensors to facilitate identification of the topology of the distribution network in real-time using Linear Discriminant Analysis (LDA) and Regularized Diagonal Quadratic Discriminant Analysis (RDQDA). The results show that this method can leverage noisy voltage magnitude readings from load buses to accurately identify distribution system reconfiguration between radial topologies during operation under changing loads.

Distribution system model accuracy is increasingly important and using advanced metering infrastructure (AMI) data to algorithmically identify and correct errors can dramatically reduce the time required to correct errors in the models. This work proposes a data-driven, physics-based approach for grouping residential meters downstream of the same service transformer. The proposed method involves a two-stage approach that first uses correlation coefficient analysis to identify transformers with errors in their customer grouping then applies a second stage, using a linear regression formulation, to correct the errors. This method achieved >99% accuracy in transformer groupings, demonstrated using EPRI's Ckt 5 model containing 1379 customers and 591 transformers.

The use of grid-edge sensing in distribution model calibration is a significant aid in reducing the time and cost associated with finding and correcting errors in the models. This work proposes a novel method for the phase identification task employing correlation coefficients on residential advanced metering infrastructure (AMI) combined with additional sensors on the medium-voltage distribution system to enable utilities to effectively calibrate the phase classification in distribution system models algorithmically. The proposed method was tested on a real utility feeder of ∼800 customers that includes 15-min voltage measurements on each phase from IntelliRupters® and 15-min AMI voltage measurements from all customers. The proposed method is compared with a standard phase identification method using voltage correlations with the substation and shows significantly improved results. The final phase predictions were verified to be correct in the field by the utility company.

The rapid increase in penetration of distributed energy resources on the electric power distribution system has created a need for more comprehensive interconnection modeling and impact analysis. Unlike conventional scenario-based studies, quasi-static time-series (QSTS) simulations can realistically model time-dependent voltage controllers and the diversity of potential impacts that can occur at different times of year. However, to accurately model a distribution system with all its controllable devices, a yearlong simulation at 1-second resolution is often required, which could take conventional computers a computational time of 10 to 120 hours when an actual unbalanced distribution feeder is modeled. This computational burden is a clear limitation to the adoption of QSTS simulations in interconnection studies and for determining optimal control solutions for utility operations. The solutions we developed include accurate and computationally efficient QSTS methods that could be implemented in existing open-source and commercial software used by utilities and the development of methods to create high-resolution proxy data sets. This project demonstrated multiple pathways for speeding up the QSTS computation using new and innovative methods for advanced time-series analysis, faster power flow solvers, parallel processing of power flow solutions and circuit reduction. The target performance level for this project was achieved with year-long high-resolution time series solutions run in less than 5 minutes within an acceptable error.

This paper discusses a solar photovoltaic (PV) DC microgrid system consisting of a PV array, a battery, DC-DC converters, and a load, where all these elements are simulated in MATLAB/Simulink environment. The design and testing entail the functions of a boost converter and a bidirectional converter and how they work together to maintain stable control of the DC bus voltage and its energy management. Furthermore, the boost converter operates under Maximum Power Point Tracking (MPPT) settings to maximize the power that the PV array can output. The control algorithm can successfully maintain the output power of the PV array at its maximum point and can respond well to changes in input irradiance. This is shown in detail in the results section.

This paper discusses the findings from an EPRI industry survey mapping the state, current practices, and future needs of distribution load modeling in the U.S. and internationally. The paper provides a benchmark for distribution utilities and a view of the current industry state and future needs for researchers and other readers. The survey found the parameters and measurements available and utilized for load modeling to vary widely between the utilities and data types. Loads were found to be largely modeled based on different load allocation methods. While distribution planning was found to focus on assessing peak load conditions, some utilities evaluate other time instances and/or explore time-series assessments. Simple grid edge and voltage sensitivity models were found common. The identified future needs include access for additional data, as well as methods to process and utilize the increasing data, handle masked load, and perform time-series load modeling.

Distribution systems with high levels of solar PV may experience notable changes due to external conditions, such as temperature or solar irradiation. Fault detection methods must be developed in order to support these changes of conditions. This paper develops a method for fast detection, location, and classification of faults in a system with a high level of solar PV. The method uses the Continuous Wavelet Transform (CWT) technique to detect the traveling waves produced by fault events. The CWT coefficients of the current waveform at the traveling wave arrival time provide a fingerprint that is characteristic of each fault type and location. Two Convolutional Neural Networks are trained to classify any new fault event. The method relays of several protection devices and doesn't require communication between them. The results show that for multiple fault scenarios and solar PV conditions, high accuracy for both location and type classification can be obtained.

By strategically curtailing active power and providing reactive power support, photovoltaic (PV) systems with advanced inverters can mitigate voltage and thermal violations in distribution networks. Quasi-static time-series (QSTS) simulations are increasingly being utilized to study the implementation of these inverter functions as alternatives to traditional circuit upgrades. However, QSTS analyses can yield significantly different results based on the availability and resolution of input data and other modeling considerations. In this paper, we quantified the uncertainty of QSTS-based curtailment evaluations for two different grid-support functions (autonomous Volt-Var and centralized PV curtailment for preventing reverse power conditions) through extensive sensitivity analyses and hardware testing. We found that Volt-Var curtailment evaluations were most sensitive to poor inverter convergence (-56.4%), PV time-series data (-18.4% to +16.5%), QSTS resolution (-15.7%), and inverter modeling uncertainty (+14.7%), while the centralized control case was most sensitive to load modeling (-26.5% to +21.4%) and PV time-series data (-6.0% to +12.4%). These findings provide valuable insights for improving the reliability and accuracy of QSTS analyses for evaluating curtailment and other PV impact studies.

Advanced solar PV inverter control settings may not be reported to utilities or may be changed without notice. This paper develops an estimation method for determining a fixed power factor control setting of a behind-the-meter (BTM) solar PV smart inverter. The estimation is achieved using linear regression methods with historical net load advanced metering infrastructure (AMI) data. Notably, the BTM PV power factor setting may be unknown or uncertain to a distribution engineer, and cannot be trivially estimated from the historical AMI data due to the influence of the native load on the measurements. To solve this, we use a simple percentile-based approach for filtering the measurements. A physics-based linear sensitivity model is then used to determine the fixed power factor control setting from the sensitivity in the complex power plane. This sensitivity parameter characterizes the control setting hidden in the aggregate data. We compare several loss functions, and verify the models developed by conducting experiments on 250 datasets based on real smart meter data. The data are augmented with synthetic quasi-static-timeseries (QSTS) simulations of BTM PV that simulate utility-observed aggregate measurements at the load. The simulations demonstrate the reactive power sensitivity of a BTM PV smart inverter can be recovered efficiently from the net load data after applying the filtering approach.

Due to the existence of DC-DC converters, fast-tripping fault location in DC power systems is of particular importance to ensure the reliable operation of DC systems. Traveling wave (TW) protection is one of the promising approaches to accommodate fast detection and location of faults in DC systems. This paper proposes a numerical approach for a DC system fault location using the concept of TWs. The proposed approach is based on multiresolution analysis to calculate the TW signal's wavelet coefficients for different frequency ranges, and then, the Parseval theorem is used to calculate the energy of wavelet coefficients. A curve-fitting approach is used to find the best curve that fits the Parseval energy as a function of fault location for a set of curve-fitting datapoints. The identified Parseval energy curves are then utilized to estimate the fault location when a new fault is applied on a DC cable. A DC test system simulated in PSCAD/EMTDC is used to verify the performance of the proposed fault location algorithm.

As a result of the increase in penetration of inverter-based generation such as wind and solar, the dynamics of the grid are being modified. These modifications may threaten the stability of the power system since the dynamics of these devices are completely different from those of rotating generators. Protection schemes need to evolve with the changes in the grid to successfully deliver their objectives of maintaining safe and reliable grid operations. This paper explores the theory of traveling waves and how they can be used to enable fast protection mechanisms. It surveys a list of signal processing methods to extract information on power system signals following a disturbance. The paper also presents a literature review of traveling wave-based protection methods at the transmission and distribution levels of the grid and for AC and DC configurations. The paper then discusses simulations tools to help design and implement protection schemes. A discussion of the anticipated evolution of protection mechanisms with the challenges facing the grid is also presented.

This paper presents a new method for detecting power quality disturbances, such as faults. The method is based on the dynamic mode decomposition (DMD)-a data-driven method to estimate linear dynamics whose eigenvalues and eigenvectors approximate those of the Koopman operator. The proposed method uses the real part of the main eigenvalue estimated by the DMD as the key indicator that a power quality event has occurred. The paper shows how the proposed method can be used to detect events using current and voltage signals to distinguish different faults. Because the proposed method is window-based, the effect that the window size has on the performance of the approach is analyzed. In addition, a study on the effect that noise has on the proposed approach is presented.

The rising penetration levels of photovoltaic (PV) systems within distribution networks has driven considerable interest in the implementation of advanced inverter functions, like autonomous Volt- Var, to provide grid support in response to adverse conditions. Quasi-static time-series (QSTS) analyses are increasingly being utilized to evaluate advanced inverter functions on their potential benefits to the grid and to quantify the magnitude of PV power curtailment they may induce. However, these analyses require additional modeling efforts to appropriately capture the time-varying behavior of circuit elements like loads and PV systems. The contribution of this paper is to study QSTS-based curtailment evaluations with different load allocation and PV modeling practices under a variety of assumptions and data limitations. A total of 24 combinations of PV and load modeling scenarios were tested on a realistic test circuit with 1,379 loads and 701 PV systems. The results revealed that the average annual curtailment varied from the baseline value of 0.47% by an absolute difference of +0.55% to -0.43 % based on the modeling scenario.

Abstract not provided.

Abstract not provided.

Networked microgrids are clusters of geographically-close, islanded microgrids that can function as a single, aggregate island. This flexibility enables customer-level resilience and reliability improvements during extreme event outages and also reduces utility costs during normal grid operations. To achieve this cohesive operation, microgrid controllers and external connections (including advanced communication protocols, protocol translators, and/or internet connection) are needed. However, these advancements also increase the vulnerability landscape of networked microgrids, and significant consequences could arise during networked operation, increasing cascading impact. To address these issues, this report seeks to understand the unique components, functions, and communications within networked microgrids and what cybersecurity solutions can be implemented and what solutions need to be developed. A literature review of microgrid cybersecurity research is provided and a gap analysis of what is additionally needed for securing networked microgrids is performed. Relevant cyber hygiene and best practices to implement are provided, as well as ideas on how cybersecurity can be integrated into networked microgrid design. Lastly, future directions of networked microgrid cybersecurity R&D are provided to inform next steps.

Secondary networks are used to supply power to loads that require very high reliability and are in use around the world, particularly in dense urban downtown areas. The protection of secondary network systems poses unique challenges. The addition of distributed energy resources (DERs) to secondary networks can compound these challenges, and deployment of microgrids on secondary networks will create a new set of challenges and opportunities. This report discusses secondary networks and their protection, the challenges associated with interconnecting DERs to a secondary network, issues expected to be associated with creating microgrids on secondary networks, standards that deal with these challenges and issues, and suggestions for research and development foci that would yield new means for addressing these challenges.

With inverter-based distributed energy resources (DERs) becoming more prevalent in grid-connected or islanded distribution feeders, a better understanding of the performance of these devices is needed. Increasing the amount of inverter-based generation, and therefore reducing conventional generation, i.e. rotating machines and synchronous generators, decreases generation sources with well-known characteristic responses for unbalanced and transient fault conditions. This paper experimentally tests the performance of commercial grid-forming inverters under fault and unbalanced conditions and provides a comparison between grid-forming inverters and their grid-following counterparts.

The benefits and risks associated with Volt-Var Curve (VVC) control for management of voltages in electric feeders with distributed, roof-top photovoltaic (PV) can be defined using a stochastic hosting capacity analysis methodology. Although past work showed that a PV inverter's reactive power can improve grid voltages for large PV installations, this study adds to the past research by evaluating the control method's impact (both good and bad) when deployed throughout the feeder within small, distributed PV systems. The stochastic hosting capacity simulation effort iterated through hundreds of load and PV generation scenarios and various control types. The simulations also tested the impact of VVCs with tampered settings to understand the potential risks associated with a cyber-attack on all of the PV inverters scattered throughout a feeder. The simulation effort found that the VVC can have an insignificant role in managing the voltage when deployed in distributed roof-top PV inverters. This type of integration strategy will result in little to no harm when subjected to a successful cyber-attack that alters the VVC settings.

Abstract not provided.

Distribution system models play a critical role in the modern grid, driving distributed energy resource integration through hosting capacity analysis and providing insight into critical areas of interest such as grid resilience and stability. Thus, the ability to validate and improve existing distribution system models is also critical. This work presents a method for identifying service transformers which contain errors in specifying the customers connected to the low-voltage side of that transformer. Pairwise correlation coefficients of the smart meter voltage time series are used to detect when a customer is not in the transformer grouping that is specified in the model. The proposed method is demonstrated both on synthetic data as well as a real utility feeder, and it successfully identifies errors in the transformer labeling in both datasets.

Distributed photovoltaic (PV) systems equipped with advanced inverters can control real and reactive power output based on grid and atmospheric conditions. The Volt-Var control method allows inverters to regulate local grid voltages by producing or consuming reactive power. Based on their power ratings, the inverters may need to curtail real power to meet the reactive power requirements, which decreases their total energy production. To evaluate the expected curtailment associated with Volt-Var control, yearlong quasi-static time-series (QSTS) simulations were conducted on a realistic distribution feeder under a variety of PV system design considerations. Overall, this paper found that the amount of curtailed energy is low (< 0.55%) compared to the total PV energy production in a year but is affected by several PV system design considerations.

Calibrating distribution system models to aid in the accuracy of simulations such as hosting capacity analysis is increasingly important in the pursuit of the goal of integrating more distributed energy resources. The recent availability of smart meter data is enabling the use of machine learning tools to automatically achieve model calibration tasks. This research focuses on applying machine learning to the phase identification task, using a co-association matrix-based, ensemble spectral clustering approach. The proposed method leverages voltage time series from smart meters and does not require existing or accurate phase labels. This work demonstrates the success of the proposed method on both synthetic and real data, surpassing the accuracy of other phase identification research.

As more renewable generation connects to distribution systems, it is imminent that existing distribution feeders will be converted to microgrids-systems that offer resilience by providing the flexibility of supporting the grid in normal operation and operating as self-sustained islands when the grid is disconnected. However, inverter control and feeder protection will need to be tuned to the operating modes of the microgrid. This paper offers an insight into the issues involved by taking a case study of a real-world feeder located in the southwestern US that was converted to a microgrid with three solar PV units connecting to the feeder. Different inverter control configurations and adaptive protection using different settings for different operating conditions are proposed for safe operation of this microgrid. The solution also helps to create a framework for protection and coordination of other similar microgrids.

During the last decade, utility companies around the world have experienced a significant increase in the occurrences of either planned or unplanned blackouts, and microgrids have emerged as a viable solution to improve grid resiliency and robustness. Recently, power converters with grid-forming capabilities have attracted interest from researchers and utilities as keystone devices enabling modern microgrid architectures. Therefore, proper and thorough testing of Grid-Forming Inverters (GFMIs) is crucial to understand their dynamics and limitations before they are deployed. The use of closed-loop real-time Power Hardware-in-the-Loop (PHIL) simulations will facilitate the testing of GFMIs using a digital twin of the power system under various contingency scenarios within a controlled environment. So far, lower to medium scale commercially available GFMIs are difficult to interface into PHIL simulations because of their lack of a synchronization mechanism that allows a smooth and stable interconnection with a voltage source such as a power amplifier. Under this scenario, the use of the well-known Ideal Transformer Method to create a PHIL setup can lead to catastrophic damages of the GFMI. This paper addresses a simple but novel method to interface commercially available GFMIs into a PHIL testbed. Experimental results showed that the proposed method is stable and accurate under standalone operation with abrupt (step) load-changing dynamics, followed by the corresponding steady state behavior. Such results were validated against the dynamics of the GFMI connected to a linear load bank.