GEARED student posters session detail

Read each of our poster sessions abstracts below. Interested in contacting the students / teams? No problem. That is available as well. For your convenience, the poster sessions are organized in the following groups:

Cyber security
DER control
Distributed automation and distribution modeling
Energy storage
Climate/environmental (climate action plan)
Microgrids
Power systems engineering
Renewables generation and integration
Smart cities (smart homes, electric vehicles, etc.)
Smart inverters and power electronics

Cyber security

A machine learning approach for false data attack detection in distribution SCADA networks

Hamdi Al-Bunashee, Chris Farnell, Roy McCann – University of Arkansas

The growth of distributed generation has resulting in larger portions of electric power generation capacity being produced at medium voltage (MV) distribution circuits. Consequently, supervisory control and data acquisition (SCADA) systems have an increasing important role in the coordination of distributed generation with conventional generation. Consequently, maintaining secure SCADA networks is essential to the reliable operation of modern electric power distribution systems. This research develops a new method of detecting and protecting against false-data attacks in SCADA networks by the use of machine learning techniques. The method uses a simplified state-estimator (SE) suitable for distribution feeders using data available through existing SCADA systems. The SE values are compared to the patterns observed from SCADA reports to derive a sequence of residuals that goes to a pre-processor and segmentation algorithm to isolate the characteristics suitable for a machine-learning algorithm. Relevant features are extracted from the segmented data and then classified by the machine learning (neural network) processor. Classification results indicating a false data attack would be identified with an alert being sent to system operators. Experimental results from a SCADA hardware-in-the-loop testbed are presented. Laboratory results confirm the benefits of the proposed method.

Attach prevention strategy in microgrids via reachability 

Paprapee Buason1, Max Liu2, Hyungjin Choi1, Alfonso Valdes2 – 1) University of Illinois at Urbana-Champaign and 2) Information Trust Institute

Microgrids are increasingly being adopted in electric power distribution systems to increase resiliency and to integrate distributed energy resources, including renewable sources (e.g., wind and solar). Modern microgrids rely on complex communications and controls to ensure the stable operation. These communication infrastructures are vulnerable to attacks; thus, attacks may potentially cause the system unstable. In this work, we present the attack prevention algorithm so called “reachability analysis”. Reachability analysis determines whether a potential malicious attack can result in actions that make system an unstable state reachable from the current state. Our approach estimates all possible trajectories caused by detectable attacks ahead of time. To that end, we propose a sensitivity analysis that assesses the worst-case bounds on the impact of attack scenarios of interest while identifying reachable unstable states in the required time budget. This method can be used to protect DERs and loads under adverse conditions. Numerical tests are demonstrated in MATLAB Simulink to validate the effectiveness of our approach.

DER control

Coordinated voltage control in distribution systems with distributed generations

Alvi Newaz, Juan Ospina, M. Omar Faruque – Florida State University

This poster presents a centralized coordinated voltage control algorithm for distribution systems with distributed generations (DGs).  The control algorithm coordinates the use of available distributed generation (DG) with the traditional voltage regulation devices, taking into consideration the reactive power limits of the DG inverters to maintain the system voltage within predefined limits. With the help of the developed algorithm the reactive power generation capacity of the DG inverters can be used to cope with the dynamic generation of the DGs without increasing the operations of the traditional voltage regulation devices available.

Demand dispatch: a case study

Neil Cammardella1, Chase Lansdale2, Larry Lassetter2, Hala Ballouz2, Sean Meyn1- 1) University of Florida and 2) Electric Power Engineers

Demand dispatch is an emerging science for controlling flexible loads to provide grid services. With the proper design, a distributed control approach can simultaneously meet the needs of the power grid and its consumers. This report explains how a demand dispatch analysis is conducted and provides results from simulations performed on a real system. Distribution system analysis was performed using Grid-i, a software platform developed by Electric Power Engineers, Inc. Our analysis suggests that implementing demand dispatch on just a single feeder can lead to savings in excess of $420,000 over ten years.

Distributed energy resource aggregation system for ancillary servicesy

Tylor Slay - Portland State University

Increasing penetration of non-dispatchable renewable energy resources and greater peak-power demand present growing challenges to Bulk Power System (BPS) reliability and resilience. This research investigates the use of an Internet of Things (IoT) framework for large scale Distributed Energy Resource (DER) aggregation and control to reduce energy imbalance caused by stochastic renewable generation. The aggregator developed for this research is Distributed Energy Resource Aggregation System (DERAS). DERAS comprises two AllJoyn applications written in C++. DERAS aggregates, simulates, and controls connected DERs. The second application is the Distributed Control System (DCS), which is the interface between AllJoyn and the physical DER. DERAS used the regional transmission organization PJM’s regulation control signals, RegA and RegD, to determine DERAS performance metrics. Their regulation control signals direct power resources to negate load and generation imbalances within the BPS. The regulation capability of aggregated DER was measured using PJM’s resource performance assessment criteria. We found the use of an IoT framework for DER aggregation and control to be inadequate in the current network implementation. However, the emulated models and aggregation response to the regulated control signal demonstrates an excellent opportunity for DER to benefit the BPS.

Effect of distributed energy resources (DER) on transmission system

Mohammad Mehdi Rezvani, Shahab Mehraeen – Lousiana State University

In this paper, the effect of distributed energy resources (DER), here solar farms, on the transmission system, such as reverse power flow and change in tap of substation transformer due to reactive power control and output power of solar farm, has been studied. In order to have a better understanding of results, two solar farms with different capacities, one 5MW and the other 20MW (equal to the rated power of a substation transformer), is used.  Moreover, this study tries to define a way for co-simulation of distribution and transmission system by proposing a universal distribution + transmission load flow, which can be used for future works. This universal load flow is conducted by simultaneous simulation of the transmission system and distribution grid in MATLAB and OpenDSS software, respectively.  Here, a part of Entergy transmission network and distribution grid, which is connected to one of transmission substation, is used as a tested network. In addition, the daily solar output power is based on the real data, which is recorded by the rooftop solar panel at ECE department_LSU.  Results of this study will be given to the Entergy/Transmission group for their future work.

Optimal energy management of microgrid using sampling-based model predictive control

Juan Ospina, Mario Harper, Alvi Newaz, M. Omar Faruque, Rick Meeker – Florida State University

This research presents a control solution capable of handling the intermittent nature of PV power generation by controlling an energy storage system and an HVAC responsive load under a real-time price scheme. The proposed controller controls the charging and discharging operations of an energy storage (ES) system and the duty cycle of an HVAC system to minimize the overall energy costs. The model uses PV and load forecasts, together with a real-time price signal, to generate a graph of control actions that can be traversed to find the optimal actions the system needs to take to minimize costs. Simulations are presented to compare the proposed method with two baseline power control schemes based on predefined charge and discharge operations and a hysteresis control of the HVAC. Simulations show that the proposed control solution achieves substantial cost savings when compared with the baselines.

Distributed automation and distribution modeling

Distributed control vs. central control of grid

Andrew Meng, Son Nguyen, Justin Tran, Qiyuan Wu - University of California, San Diego

For a very long time since the inception of the power grid, the design of the power grid has remained relatively the same. The system revolved around a centralized system of electricity generation, storage, and distribution. This system has provided sufficient energy for over a century, but recent growth in demand for power and energy has caused a shift to a newer and more efficient distributed energy grid. In a distributed energy grid, smaller power generation and storage systems are spread out across the country and used to power communities. In this paper, we will focus on different features and applications of a distributed power grid and its significance in our modern day system. This paper will provide a comprehensive overview of a distributed energy system and its benefits as compared to a centralized energy system.

Meter and device placement for Duke Energy

Shelby Tomassi, Andrew Simms, Ajani Nisbett, Matthew Tauber - University of North Carolina at Charlotte

Research shall be conducted on Duke’s Distribution Management System (DMS) software in the Duke Energy Smart Grid Laboratory. The project first covers the placing of voltage meters and later lead into the placing of Integrated Volt/Var Control (IVVC) devices using the results of meter placement. The process of meter placement will be carried out by conducting sensitivity analysis on switches across a Duke Energy circuit in DMS. This analysis will provide insight as to which switch placements are more sensitive than others within the circuit. The switches that will undergo the first sensitivity analysis are external switches in the circuit. To examine the 4-3-2 rule, nodes will be examined in the sensitivity analysis as switches will not always be located are areas that correspond with the 4-3-2 rule. This rule states that for every 400 customers, 3 miles of line, or 2 megawatts of power, a device must be placed. A device should be placed for whichever of these three criteria is met first.

The Role of Electric Vehicle for Frequency Regulation during Grid Restoration

Towfiq Rahman and Dr. Zhihua Qu - University of Central Florida

Power system restoration after a blackout involves generation restart, energization of buses/power lines and sequential load pickup. During this process, it is of paramount importance to maintain system frequency stable to prevent system collapse or cascading events. As a fast response energy resources in smart grid with bidirectional power flow capability, Electric Vehicle (EV) can be used to compensate for load-generation imbalance in the system which stabilizes frequency and accelerates the load picks up the process. In this paper, an optimization problem is formulated which utilizes EVs to provide for frequency regulation over the restoration time horizon. At the end of restoration, the EVs are also returned to their initial State of Charge (SOC). The problem is solved as a finite time-optimal control problem which determines the global optimal charging/discharging trajectories of the EV batteries which regulates system frequency and also returns the SOC of the batteries to the original state at the end of restoration. Simulations are carried out to illustrate the analytical results.

Smart inverter functions impact on dynamic regulator settings for distribution voltage control

H M Mesbah Maruf, Dr. Badrul Chowdhury - University of North Carolina at Charlotte

Dynamic step voltage regulator settings for voltage control scheme is proposed in an active distribution system in the presence of distributed photovoltaic (PV) systems to mitigate excessive tap operations during high PV variability period. Voltage regulating devices, such as load tap changers, step voltage regulators, and capacitor banks need to be properly controlled and coordinated while interacting with the smart inverter functionalities to improve power quality, and system efficiency. The proposed dynamic regulator settings use a unidirectional communication link to update its controller setpoints based on the PV and load forecast to achieve the desired voltage profile throughout the feeder. The proposed scheme is tested on a modified IEEE test system with PV integrated using OpenDSS and Matlab COM interface and the performance under different system operation condition is evaluated.

Energy storage

Battery testing according to the new EPRI guide and applications to distribution systems

Oluwaseun M. Akeyo, Vandana Rallabandi, Nicholas Jewell , Dan M. Ionel  - University of Kentucky

This work reviews the procedures, layouts and metrics described in the new test manual issued by the Electric Power Research Institute (EPRI), in order to determine the performance and functionality of utility and distribution scale battery energy storage systems (BESS). In this approach, the large battery unit is connected to the dc-link of a bidirectional power conversion system (PCS), which may be interfaced with either the utility grid or a load bank. The BESS is tested at different charge and discharge cycles in order to estimate its operation and performance characteristics, some of which include; available charge/discharge energy at rated power, continuous charge/discharge duration, battery ramp rate, and ac round trip efficiency (RTE). Furthermore, tests were proposed and conducted for determining the equivalent circuit parameters of large batteries based on which a runtime battery unit equivalent circuit model was developed. Its application for grid services was demonstrated through an example primary frequency response. For the purpose of validating the equivalent BESS model, experimental results retrieved from the LG&E and KU E.W. Brown solar facility, which houses a 1MW/2MWh operational BESS and a 1MVA variable load bank were compared with simulation results from the equivalent model developed in PSCAD/EMTDC software.

Chemically based utility scale battery energy storage

Nasser Althaiban, Philip Murphy, Madison Wynn - University of North Carolina at Charlotte

The poster is for designing a chemical battery utility scale storage system that can be utilized in between the grid and solar array. This battery energy storage system (BESS) will be expected to be economically viable while  helping to mitigate many of the existing issues located on the grid such as voltage sag & flicker, solar intermittency, PV firming, energy time-shift, and peak demand reduction. The poster explains the process of picking the chemistry of the energy storage and shows how is the OpenDSS software is going to be used for deciding where to locate the BESS.

Electrical energy grid storage system technologies

Jessica Chen, Brian Wong, Antonio Segura, Kevin Kao – University of California – San Diego

Electrical Energy Storage (EES) is a system of energy storage that converts electrical energy into a form of energy that can be stored for later use. It is used to provide additional energy as the supply and demand for energy from the grid vary throughout the course of the day. This paper will explore the following four configurations of EES: flywheel, pumped-hydroelectric, compressed air, and battery storage. An overview of each energy storage system is provided. The costs involved to operate the storage system, ability to efficiently store and deliver energy, and hindrances are analyzed to discuss the practicality of each system.

Climate/environmental (climate action plan)

Line sagging in adverse condition

Gun Park – University of California, San Diego

This poster deals with line sagging in adverse conditions. As the temperature drop during the winter season increased, increasing number of transmission towers broke due to shortening of the conductors.

Microgrids

A method for non-intrusive power consumption monitoring using giant magnetoresistive sensors

Shane Witsell, Jonathan Hammer, Yusuf Ozturk – San Diego State University

With accelerating Smart Grid and automated Smart Building technologies, there is an ever-increasing demand for non-invasive, networked load monitoring.  Magnetometers leveraging Giant Magnetoresistance (GMR) can be used to measure electrical currents via their inherent magnetic fields.  Research is conducted to characterize these sensors and investigate their usefulness in a wireless embedded system for the purposes of energy management in power distribution equipment The goal of such a system is that it can be installed without opening the electrical equipment or removing the faceplate, so that power management can be employed without requiring qualified electricians. This work investigates realistic scenarios for employing GMR sensors in power distribution load centers with multiple circuits. Research is done to assess how multiple current sources create magnetic fields in panelboards, and how an array of GMR-based magnetic field sensors may be used to capture power load events. Methods for collecting and processing multi-channel sensor data are discussed, particularly techniques for eliminating signal crosstalk between the channels. Through several experiments on two purpose-built testbeds, a calibration method is developed for the sensor array. Using the calibrated sensor array, measurements of electrical currents are made with a percent error ranging from 0.82% to 2.5% for active loads.

Condition monitoring of a campus microgrid for optimum power delivery

Mazen Aljohaif1, Ali Eskandrany1, Waleed Abumelha1, Fawaz Aljahdali1 and Kaisar R. Khan1,2 - 1) Department of Electrical Engineering and Computer Science and 2) Entergy Louisiana

A Microgrid (MG) is an independent grid with both load and its own generation that has the option to be connected to the grid or can be islanded. Intelligent monitoring and control of MG make it as a building block for smart grid technology. With automation and distributed control, load and generation management can be done remotely. A distributed control system (DCS) with advance smart sensor network have been deployed to control the energy delivery system of an MG established in a university campus. As shown in figure 1 this MG has high efficiency combined heat and power (CHP) generation and photovoltaic energy sources. We conduct a comparative study of the effects of Condition Monitoring and Intelligent Control (CMIC) on a Micro-grid (MG) using distributed control and wireless technology. Results shows that, by implementing condition monitoring based automation, an economic, efficient and reliable energy delivery system can be achieved.

Designing a reliable microgrid for the power grid of Our Lady of the Lake (OLOL) regional medical center

Mohammad Mehdi Rezvani, Shahab Mehraeen – Louisiana State University

The goal of this paper is to find out the extra power generation that should be added to the existing generation capacity of the power grid of Our Lady of the Lake (OLOL) Regional Medical Center in order to have reliable operation during the islanded mode, when it is disconnected from Entergy feeders in a time of flood or etc.  In this project, first, the existing power grid of OLOL and some changes that will be added to it in the near future, such as building new children hospital and installing new chillers, will be modeled in MATLAB/Simulink 2017b. Then based on this model, the extra generation that can satisfy new total loads and add some reserve to its power grid will be determined. Subsequently, the size of the new extra generator and change in capacity of existing generators by replacing one of them with higher power generator will be specified. Finally, results will be given to the Entergy/Distribution group to implement.

Enhancement of transactive energy test bed as related to microgrid with deployment of synchrophasors for protection, monitoring and control

Berthony Lozier, Aiden Krombach, Weng Yu - SUNY Buffalo State College

The growing need for clean renewable power sources led to the increased deployment of distributed generation (DG) resources in the electrical grid. While this provides numerous benefits, it also requires higher level techniques for controlling, monitoring, and dispatching the dispersed assets. The concept of Transactive Energy made its way to the forefront of discussions regarding the management of the modern grid.    The project provides means for demonstrating and testing a new set of techniques to oversee the operation of the microgrid using a testbed consisting of combination of microturbine, wind generation, photovoltaic systems, and battery storage.     A Supervisory controls and data acquisition (SCADA) system provides monitoring and control of each of the components. Various Schweitzer Engineering Laboratories (SEL) relays are used to protect the system as well as to provide synchrophasor measurement data to a Human Machine Interface (HMI) for evaluation of system stability and performance.  Battery management systems was integrated into SCADA to monitor energy storage. The testbed enables sharing of electric power in Transactive Energy environment.

Hierarchical optimal control of potsdam microgrid in islanded mode

Jimiao Zhang1, Jie Li1, Lei Wu2 – 1) Clarkston University and 2) Stevens Institute of Technology

This project studies a three-level hierarchical control strategy for the resilient community underground microgrid in Potsdam, NY, operated in islanded mode. Specifically, an AC Optimal Power Flow (ACOPF) problem is solved via rank-relaxed semidefinite programming (SDP) method on the tertiary level to derive an optimal operating point of the microgrid.  Then two decentralized secondary control methods based on proportional-integral (PI) control are proposed; one is to make PI controllers continually act, and the other is to periodically reset the PI outputs to zero. Both methods allow distributed generation (DG) units to closely track the optimal economic dispatch commands by generating frequency correction terms to the P – f droop controllers. They are easy to implement and guarantee fast tracking with small steady state error. This paper studies a three-level hierarchical control strategy for the resilient community underground microgrid in Potsdam, NY, operated in the islanded mode. Specifically, an AC Optimal Power Flow (ACOPF) problem is solved via rank-relaxed semidefinite programming (SDP) method on the tertiary level to derive an optimal operating point of the microgrid.  Then two decentralized secondary control methods based on proportional-integral (PI) control are proposed; one is to make PI controllers continually act, and the other is to periodically reset the PI outputs to zero. Both methods allow distributed generation (DG) units to closely track the optimal economic dispatch commands by generating frequency correction terms to the P – f droop controllers. They are easy to implement and guarantee fast tracking with small steady state error. Simulation studies are conducted with PSCAD/EMTDC, where the proposed methods are compared.

Machine learning for critical reliability improvement during disasters and prolonged outages

Lizon Maharjan - University of Texas at Dallas

Current power distribution system faces challenges that originate from aging infrastructure, DER integration, and frequent natural disasters. An ideal solution will include a retro-fit device that can utilize existing infrastructure and integrate DERs advantageously. One of such solutions have been visualized using Advanced Micro Grids. The presented study utilizes Machine learning algorithms to create energy management systems for such micro-grids with the goal of maximizing power availability to the critical loads of the community. The residential level electronics with integrated artificial intelligence have been designed, developed, and tested as a part of this effort.

OpenFMB enabled power system operations using controller hardware-in-loop simulation

Rasik Sarup, Madhav Manjrekar - University of North Carolina, Charlotte

The Department of Energy defines Microgrid as a group of interconnected loads and distributed energy resources within clearly defined electrical boundaries that acts as a single controllable entity with respect to the grid. A microgrid can connect and disconnect from the grid to operate in grid-connected or island mode. This application of microgrid is of key importance in the real-time simulation performed for simulated feeder system. The problem statement is to simulate a microgrid connected distribution feeder model in controller hardware in loop and implement use-cases for 61850-GOOSE protocol for the proposed architecture using OpenFMB framework. The idea to configure analog inputs and outputs of simulation in Protection and Control (P&C) Intelligent Electronic Device (IED) and communicate data to RTU using 61850-GOOSE protocol is researched and implemented. The real work is embedded in converting RTU’s output data into respective understandable language for HIL’s subscriber and publisher block. To perform this operation, a new North American Energy Standard interoperability framework known as Open Field Message Bus (OpenFMB) is used. The validation of protection and performance of real-time grid following and forming mode in a microgrid with 800kW of Distributed Energy Resources is the catch of this detailed research work.

Optimal scheduling of MG in a joint energy and AS market

Yiwei Wu, Jian Shi, Gino Lim - University of Houston

In this poster, we present a novel microgrid scheduling strategy considering the ancillary service preparation and implementation into a co-optimization framework. Due to the large scale uncontrollable renewable energy in microgrid which requires frequency regulation reserve, and exist of operation mode switching scenarios which are involved with alternation of ancillary service provision and different reserve, we provide a set of Islanding rules as a more comprehensive approach to cover all states of MG operation with islanding events.  Especially, a combination method of power exchange for frequency control market approach and self-sustain approach is adopted to overcome the operation model switching uncertainty. A two-stage microgrid scheduling approach is proposed based on chance-constrained programming that allows the microgrid operator to determine the optimal reserve amount and risk preference. The original two-stage chance-constrained model is approximated as a linear mix-integer programming model. Finally, we conduct policy study on microgrid scheduling from different aspects of data set including SOR levels, price setting, capacity setting and expected islanding event to show the application of proposed model and strategies.

Restoration using distributed energy resources for resilient power distribution networks

Ranadhir Sarkar, Dr. Zhihua Qu – University of Central Florida

In this work, we propose a restorative method to minimize the unserved demand in a power distribution network where some islands are formed due to multiple line outages. The restoration is achieved by tie-line switching and by solving a mixed-integer non-linear problem. Due to combinatorial issues of the optimization problem, a method is proposed which restores the system by setting up Microgrid boundaries within the islands and connecting them through tie-lines for exchange of active power in between them. The optimal tie-line to interconnect the Microgrids is decided through information exchange between the Microgrid operators. Case studies on IEEE 33-bus distribution network is presented to illustrate the results and validate the performance of the proposed method.

Thermal baseline study of uptown and the role of grid resiliency in reducing carbon emissions

Erick Bittenbender, Sabrina Nguyen, Dr. Katrina Kelly-Pitou, Dr. Gregory Reed - University of Pittsburgh

With more cities, states, and countries laying out emissions reduction goals, questions arise about what technical solutions are best oriented to help these entities meet their goals. One potential solution is the use of microgrids to help improve power system resiliency and promote the use of more carbon-neutral generation. By pursuing microgrid development, opportunities also exist to create localized heating networks, commonly referred to as district heating, that can also assist in reducing carbon emissions. This work explores the Uptown neighborhood of Pittsburgh, PA, including what kind of thermal demand exists in the area, what level of emissions are produced, and what potential solutions may improve this neighborhood’s emissions profile. In addition to exploring the neighborhood itself, the implications of improving system resiliency and the potential impact this may have on carbon reductions targets is also explored.

UCF campus grid modeling with MGMS and OPAL-RT

Lisian Shehu, Ivelisse Rivera, Inalvis Alvarez Fernandez, Dr. Wei Sun – University of Central Florida

This project aims to create a microgrid model of the University of Central Florida to analyze how an energy management system determines the optimal mode of operation based on generation, load, and price of energy at any specific time. Data on generation, load and price of energy data have been collected and generated to model the UCF microgrid by utilizing Information Model Management (IMM) and MATLAB Simulink. The collected data and the created model will be uploaded into the Microgrid Management System(MGMS) for simulation. Furthermore, real-time data will be generated using OPAL-RT technology, as well as communication being established between OPAL-RT, IMM, and MGMS in order to simulate the microgrid model in real-time. The simulation results will be analyzed to determine how the energy management system can determine the optimal mode of operation in an efficient manner. The results of this project will contribute to the development on microgrids, smart grid technologies, and renewable energy.

Power systems engineering

Aggregate model of single phase induction motors

Bikrant Poudel1, Rochak Shiwakoti1, Ebrahim Amiri1, Parviz Rastgoufard1, Thomas E. Field2 – 1) University of New Orleans and 2) Entergy Corporation

This poster presents the aggregate model of single phase induction motors used in residential heat pump loads. The  equivalent circuit parameters of the aggregated motor model are derived by performing no load and locked rotor tests. The proposed  model captures the loads aggregated behavior for any arbitrary number of loads at the distribution substation, which is beneficial for  utilities’ planning studies. The proposed load model is applied to the IEEE 39 bus system and subjected to a 3-phase fault to study the  motor’s recovery and potential system delayed voltage recovery. To verify the validity of the proposed load model, results are compared  with the same case study connected to individual motor loads. The simulation results indicate that the motor will recover for 15 cycle 3-  phase faults and not for 90 cycle faults.

Adaptive overcurrent relay coordination for power distributions systems with distributed generations

Ke Xu, Yuan Liao - University of Kentucky

During the operation of a modern distribution power system, some abnormal conditions may cause power outage and damage on the equipment connected to the power system. To minimize the damage, the faulted components must be identified and isolated as quickly as possible. Therefore, a reliable protective system is needed. Also, the protection system should follow the requirements of sensitivity, selectivity, reliability and speed. In the power distribution system, the overcurrent relays are widely used for protection. The core part of the protection system is the problem of overcurrent relays coordination, which is to select the appropriate settings of each overcurrent relay to make sure the backup relays take over as quickly as possible when the primary relays fail. So far, there are some existing optimum coordination methods for the coordination of overcurrent relays. However, these methods are usually based on the conditions of fixed distributed generations output and constant load in the distribution system. When there is a variation of distributed generations output or load in distribution system, these existing optimum coordination methods may not offer the most appropriate settings for overcurrent relays anymore. Therefore, an online adaptive optimum coordination of overcurrent relays is a solution to solve this problem.

Fault classification and location identification in a smart distribution network using ANN 

Muhammad Usama Usman, Juan Ospina, M. Omar Faruque – Florida State University

This paper presents a novel approach to classify and locate different types of faults in a smart distribution network (DN). The proposed method is able to classify all types of faults that can occur in a DN and then based on fault type, it can identify the approximate fault location (FL) with a high accuracy. The method is based on artificial neural networks pattern recognition which uses data from D-PMUs/smart meters placed at different locations in a DN. The proposed technique needs fault-on voltages of all the nodes connected to the end of line/branches in order to classify and locate different types of faults. The method is tested on a modified IEEE-37 bus system with distributed generation along with dynamic loading conditions and varying fault resistances. Both balanced and unbalanced fault types are applied to the system. An accurate classification of 100% is achieved when classifying all fault types and above 99% accuracy is achieved when identifying the approximate fault location.

High level investigation of nonintrusive load monitoring (NILM) 

Dekwuan Stokes, Aryana Nakhai, Ryan Brody, Adam Emes, Dr. Robert Kerestes – University of Pittsburgh

This research focused on a Non-Intrusive Load Monitoring (NILM) using MATLAB. The algorithm tested waveforms using Dranetz meters and Eaton power lab benches in the University of Pittsburgh’s power lab. The waveforms were placed into the NILM algorithm where it used the pre/post-transient data to classify whether the load is resistive, capacitive, or inductive. Within this algorithm a separate function read in the waveforms and used basic power equations and concepts in order to print out the power consumption and compare to the data collected from the Dranetz meters. This project utilizes the knowledge of power systems, signals processing and coding and could be used to improving smart grid applications and smart home technology. Also, these results will assist in creating and developing projects for undergraduate students. The research creates an opportunity where students are able to understand more of the smart grid and how different loads consume power in households.

Improving DC circuit breaker performance through an alternate commutating circuit 

Sudipta Sen1, Dr. Shahab Mehraeen1, Dr. Farzad Ferdowsi2, 1) Louisiana State University and 2) University of Louisiana at Lafayette

Fault interruption in dc circuits is more challenging than in their ac counterparts. The absence of natural current zero crossing along with resistive nature of the dc grids creates a significantly higher fault current to disrupt at the dc circuit breakers.   Available approaches to break dc fault currents include creating forced zero-crossing current at the breaker or employing solid-state circuit breakers. This paper summarizes the current dc breaker technologies and proposes a new alternative method. The proposed mechanism is a mechanical circuit breaker that utilizes two switches, of which one generates zero-crossing with an alternate oscillatory circuit for the other one, which can be a conventional zero crossing-based ac breaker and is used in the main circuit. This is different from the conventional single-switch commute-and-absorb method currently used. It is shown that the proposed oscillatory circuit improves the fault current extinction and significantly reduces the voltage rate-of-change while creating the current zero-crossing faster when compared to the available technology. Thus, the proposed mechanism is capable of interrupting high dc currents with a minimal arc through a less expensive ac circuit breaker. Simulation and hardware results are provided to show the efficiency of the proposed breaker.

Power distribution network reconfiguration considering distributed generations

Ahmad Ghaweta, Yuan Liao – University of Kentucky

Feeder reconfiguration is performed by changing the open/close status of switches. Primary distribution networks contain two types of switches, known as tie switches (normally open) and sectionalizing switches (normally closed). These switches are designed for both protection and configuration purposes. A whole feeder or part of a feeder may be served from another feeder by closing a tie switch linking the two while an appropriate sectionalizing switch must be opened to maintain the radial structure of the system. In this paper, the problem is formulated as a multi-objective problem considering four objectives related to minimization of the system power loss, minimization of the deviations of the nodes voltage, minimization of branch current violation and minimization of feeder’s currents imbalance. Since these objectives are different and difficult to be solved by the conventional approaches that may optimize only a single objective. A new approach, based on the Differential Evolution algorithm (DE), is presented, allowing the topological and electrical constraints to be satisfied. DE algorithm is a new heuristic approach mainly having three advantages: independence on the initial population values in finding the global minima, fast convergence, and using few control parameters. This paper also aims to include the effects of Distributor Generators (DGs) outputs and to find the optimal allocation and size of the planned DGs units. The validity and effectiveness of the proposed algorithm are demonstrated using two different test systems, IEEE 16 and 33-bus distribution network having three substations and one substation respectively. The proposed algorithm is also compared to some alternative methods such as Genetic Algorithm.  

Transmission line parameter estimation using PMU measurements

Mustafa Lahmar, Yuan Liao - University of Kentucky

Transmission line parameters are necessary inputs for performing all sorts of analyses and control to maintain reliable and safe operation of the power system. Mainly, protective relaying based on impedance relies on accurate line parameters. Online parameter estimation is also desirable for estimating line temperature and thermal conditions of transmission lines in real time. This paper presents several methods for estimating positive sequence parameters of transmission lines considering different transmission line configurations. The methods are based on a linear least squares technique for improved computational efficiency. Synchronized data obtained from the phasor measurement unit are assumed available for estimating line parameters, while potential synchronization error can be estimated with using proper modeling method. Based on actual applications, different transmission line models were used including long, medium, and short lines. Evaluation results have demonstrated that the proposed methods are highly accurate and efficient.

Renewables generation and integration

A new role for conventional power generators in a DER-dominated electric system

Gustavo Cuello-Polp, Efraín O'Neill-Carrillo - Univesity of Puerto Rico

This work focuses on the new supporting role of large, fossil-based generators in a power system with an increased level of renewable energy. The goal is to provide recommendations about the establishment of fair and transparent operational and regulatory criteria to ensure sustainability and resiliency in power grids with increased levels of PV-based Distributed Energy Resources.     The current regulatory framework was studied and the requirements to model jointly bulk power system and distribution systems were identified. The case study was Puerto Rico’s electric grid with different levels of solar generation. It was assumed that existing thermal power plants would continue operating albeit at a reduced output, and with new, flexible generators (e.g., capable of cycling). It is envisioned that the Puerto Rican power infrastructure could be divided into regions that can operate interconnected or stand-alone. The system has been operated like this in the past, although only during emergencies. DIgSILENT software was used for the simulations which included the 230 kV and 115 kV transmission system, the 38 kV subtransmission system and a simplified model of parts of the distribution system.

Applying H2 optimization methods for improved frequency regulation in multi-area systems with >75% wind generation

Muthanna Alsarray, Roy McCann – University of Arkansas

There is presently 80 GW of wind energy installed in the US and Southwest Power Pool reported 60% instantaneous wind penetration in March 2018. California CAISO achieved 50% instantaneous capacity from solar on March 6, 2018. These trends are expected to continue such that with generator interconnect (GI) requests in queue could likely result in >75% instantaneous power from wind & solar inverter–based sources. This paper presents a norm-based H2 control approach to improve frequency and tie-line power regulation in multi-area power systems with large amounts of solar and wind based generation. The design method is detailed in terms of necessary sensors, measurement capability, and the modeling of system disturbances.  The method is evaluated for a two-area system with a mix of turbine and inverter-based generation. A detailed simulation is used to compare the proposed control method to a conventional area-control error design. Results indicate the improved ability to meet requirements such as NERC BAL-003-1 standards during transient conditions induced by wind  energy resources.

Community-based hybrid wind and solar PV farm

Nasser Alawhali, Vandana Rallabandi, Oluwaseun M Akeyo, Dan M. Ionel – University of Kentucky

This work presents a power system configuration and control schemes for a Multi-MW wind-solar hybrid system, which includes multiple wind turbines and solar panels to support a desalination plant. The wind and solar systems are sized such that the desalination plant obtains, as far as possible, its power from the renewable sources, to minimize the dependence on the utility grid. Each section of the wind turbine system includes a variable speed permanent magnet synchronous generator connected to the grid via a back-to-back voltage source converter. The active rectifier is controlled to operate the wind system at the maximum power point. The solar plant has multiple panels connected to the grid via a three-phase converter, performing the dual functions of power transformation and maximum power point tracking. The modeling and validation of this system was performed with the PSCAD/EMTDC. The hybrid farm is capable of operation in both stand-alone and grid connected modes.

Coordinated voltage regulator control

Shahrzad Mahdavi, Aleksandar Dimitrovski - University of Central Florida

Voltage control is one of the fundamental operational requirements for power systems, which can become very challenging in presence of distributed energy resources (DERs). This poster presents a novel method for inclusion of voltage regulators (VRs) in a distributed voltage control framework for distribution networks. The goal is to coordinate and optimize the operation of voltage regulators. Of special interest are networks with high penetration of distributed energy resources, PVs in particular, which exhibit overvoltage problems. To that end, we have developed a dynamic VR model and applied the proposed coordinated control method on a modified IEEE 123 test system with very high penetration of PVs. Over Voltage Severity Index (OVSI) and Cumulative Voltage Variation Index (CVVI) have been defined and used in the algorithm to control the regulation steps of VRs. The developed method is compared against the base case with no control and the results show its effectiveness in coordinated VR control.

Distribution power system control for low-income low-cost net zero energy communities

Huangjie Gong, Oluwaseun Akeyo, Vandana Rallabandi, PhD, Donald Colliver, PhD, P.E., F-ASHRAE, Dan M. Ionel, PhD, FIEEE - University of Kentucky

The intermittent nature of solar photovoltaic (PV) causes instability to the distribution system where they are congregated. Increasing residential solar PV penetration also brings challenges such as ‘duck curve’ to the utility. Previous research works focused on communities with high PV penetration propose the use of large battery energy storage (BES) to offset the impacts of solar power variability. Other works propose demand response and PV curtailment, which might influence consumer comfort and make it inefficient for the customers, respectively. In this work, energy storage systems at each house are controlled to mitigate the ‘duck curve’ phenomenon and maximize cost benefits to the consumer without rescheduling the HVAC systems or PV curtailment. Therefore, the consumer comfort is maintained and the benefit gains from the PV generation are maximized. In addition, electricity trading is allowed among individual homes, enabling all houses to access PV sources and residential BES in the same distribution power network. Four different types of houses with and without PV and BES are modeled using BEopt and EnergyPlus in this work.  Simulation results show that the proposed control for individual homes brings economic benefits at the house level and reduces fluctuations at the power system level.

Implementation of a ternary pumped storage hydropower system on extremely high renewable penetrated U.S. western interconnection

Zerui Dong1, Robert Nelms1, Eduard Muljadi1, Jin Tan2, Mark Jacobson2 – 1) Auburn University and 2) National Renewable Energy Laboratory

In recent years, as the rapid deployment of solar and wind energy in the United States, energy storage (ES) are more and more integrated to deal with the variability caused by renewable energy. Conventional pumped storage hydropower (C-PSH), as the most commonly used and most mature solution, unable to fit future needs of a high renewable penetrated situation. Ternary pumped storage hydropower (T-PSH), as one of the advanced PSH technologies, has attracted more attention from the industry and academia owing to its ability to provide more reliable frequency ancillary services than C-PSH. By using an innovative structural design, T-PSH can use a single generator rotating in one direction with two different runners connected to the same shaft. A new operation mode named hydraulic short-circuit (HSC) mode is added to help T-PSH provide primary frequency support and regulation service while pumping. This unique design also makes T-PSH have less transition time in operation mode switching. This poster presents implementation of first T-PSH system in the Western Interconnection grid model based on the GE Positive Sequence Load Flow (PSLF) platform. The comparison between T-PSH and C-PSH under different renewable energy penetration is included to illustrate advantages of T-PSH in providing ancillary service.

Parameterization of Aggregated Distributed Energy Resources (DER_A) Model for Transmission Planning Studies

Inalvis Alvarez Fernandez, Deepak Ramasubramanian, Anish Gaikwad, Jens Boemer – University of Central Florida

With an increased number of distributed energy resources (DERs) connected to the distribution system, visibility of the response of the aggregated DER to a transmission system event can be critical to power system stability. This study focuses on parameterizing an aggregated model of dynamic DERs that includes advanced functions such as voltage ride through and dynamic voltage support that has been previously developed. In order to provide model parameters appropriate for an actual distribution system, dynamic simulations based on a detailed feeder model were executed to evaluate its dynamic voltage response with non-smart legacy DERs. Parameters obtained from detailed simulations were played-in a positive sequence simulation software to observe the response of the DER_A model. DER_A model’s behavior correlates with results from detailed simulation for the induced voltage events. Further studies ought to be performed to capture DER response in a wide arrangement of distributions systems to develop generic parameters.

Investigation of flicker impact and its mitigation strategy for tidal and river power generation

Sangwon Seo1, Jinho Kim1, Eduard Muljadi1, Shazreen Meor-Danial2, Monty Worthington3 - 1) Auburn University, 2) National Renewable Energy Laboratory, and 3) Ocean Renewable Power Company

In tidal and river hydrokinetic power generation, the fluctuating nature of water currents and the grid condition induce the fluctuating output power, and this power fluctuation may cause voltage fluctuations or flicker during the continuous operation. The mostly adopted technique to mitigate the flicker in renewable resources is the reactive power compensation (VAr Control) because it is cost-effective and efficiently mitigate flicker when a large amount of the reactive power exists. But, distribution network where has low X/R ratio causes voltage fluctuations proportional to the real power fluctuations. Thus, reactive power compensation alone is not sufficient to mitigate the flicker on that grid, so the battery energy storage system(BESS) that smooths the real power fluctuation is more efficient to mitigate flicker in the distribution network with low X/R ratio. With the various grid conditions, the feasibilities of the VAr control and the BESS system have been tested using a simplified simulation model of the tidal turbine developed in the simulation tool of PSCAD/EMTDC, and the short-term flicker severities are measured to compare the effectiveness of flicker mitigation with different techniques.

Protection coordination for assembly HVDC breakers for multi-terminal HVDC grids using wavelet transform

Bhaskar Mitra, Dr. Badrul Chowdhury – University of North Carolina, Charlotte

A major challenge in a multi-terminal HVDC grid is the detection and rapid removal of DC-side faults. Specially designed DC breakers are required for this purpose. The assembly HVDC breaker is one such design. The ASCB can interrupt faults within 3 − 5 ms, thereby protecting the converters in the HVDC link. Conventionally, the breakers are operated by using the current threshold. This paper proposes a coordinated control strategy for the ASCB to overcome the shortcomings of the traditional strategy. During faults, the energy released by traveling waves are utilized for detecting faults on the system and activating the breakers. Simultaneously, the change in current direction is also considered to avoid any false triggering of any of the breakers. This is applied locally to all the relays of the breakers to avoid any form of remote communication, which may be difficult to achieve at remote off-shore locations. A model of a three-terminal bipolar HVDC grid is developed in PSCAD/EMTDC and the simulation results verify the effectiveness of the proposed solution for preventing DC side faults for a bipolar HVDC system.

Scaling of permanent magnet synchronous machines for use in adjustable speed pumped storage hydropower

Patrick Haney, Tanner Grider, Eduard Muljadi, Robert Nelms, Robi Robichaud – Auburn University

Given the current trend in the implementation of renewable energy sources, better stability and frequency regulation for the grid is of utmost importance. Power storage that allows for fast-acting response along with a large capacity, while offering a long lifespan and reliability is in great demand. Adjustable-speed pumped-storage hydropower (AS-PSH) is the current dominating solution for these problems. However, the sizing of the machines required in this application varies depending on the characteristics of the specific sites where this technology is to be utilized. Therefore, readily-available motor-generators are not always viable. Custom designs require expertise that is in decline due to an aging workforce and the competition among alternative university-level curricula that may be more desirable to students approaching the workforce. This design report offers a method applying a MATLAB program to optimize the dimensioning of permanent magnet synchronous machines (PMSM) for use in AS-PSH.

Solar power forecasting using empirical copulas

Hossein Panamtash, Qun Zhou - University of Central Florida

The uncertainty associated with the solar power generation makes the task of solar power forecasting more challenging. To this point, recently most of the studies have focused on probabilistic forecasting instead of the deterministic point forecasting. But combining the probabilistic forecasts at lower levels like Photo Voltaic (PV) sites to shape a probabilistic forecast for the higher level (namely substation or system level) is a difficult task. Especially if we factor the requirement that probability distribution for the top level should match the distribution of the bottom levels. This work presents a new method of generating coherent probabilistic forecasts for an entire hierarchical system of PV generation. We propose the application of empirical copulas to the probabilistic forecast for the bottom levels to shape the top-level forecast. This method is compared to the top-down model which brakes the top-level forecast to form the bottom-level forecasts. While the copula method can capture the independent patterns for each location, it also has less computational complexity compared to the top-down method. The Pinball loss and Continuous Ranked Probabilistic Score (CRPS) scores are used to compare the error associated with each method and show that the copula method can effectively replace the top-down method.

Smart cities (smart homes, electric vehicles, etc.)

Design and analysis of modular axial flux switched reluctance motor

Rochak Shiwakoti , Bikrant Poudel, Ebrahim Amiri - University Of New Orleans

Axial Flux Switched Reluctance Motor (AFSRM) are among the most reliable, robust, high power density electric machines with a simple structure and a wide range of operating speed. Such motors are potentially appropriate choice for applications where a large diameter/axial length ratio is required, such as in-wheel propulsion electric vehicle. However, due to their salient structure and non-linear magnetic characteristic, such motors suffer from large torque/force ripple. There are primarily two approaches for keeping the torque ripple whiten the permissible limits; one method is to modify and optimize the geometric construction and magnetic design of the motor , while the other is to use sophisticated control techniques for driving the motor. However, the methods available in the literature are mainly developed for radial flux SRMs. This paper, presents a new design for AFSRMs to lower the torque ripple and improve the overall efficiency. The proposed design is based on a modular structure with four stator/rotor disks. To effectively distribute the rotational torque along the axial length of the motor, every two non-adjacent stator disks (stator 1 & 3, stator 2 & 4) are excited synchronously. The motor performance is analyzed via FEA and compared with the conventional single disk AFSRM.

Low-power bluetooth motion sensor for spinal kinematics

Austin Wolf, Yusuf Ozturk - San Diego State University

A system was designed to collect orientation data from orientation data of the spine and send data back to a mobile application. The sensor uses a low-power motion sensor that collects gyro, accelerometer and magnometer data, and fuses the data into an accurate quaternion orientation value that can be sent to a central device. This work focused on producing a sensor that could operate at low-power during sampling. The power consumption of a sensor node was tested, demonstrating 28 hour battery life.

MIST: a machine intelligent sensor topology fog computing architecture to mitigate the propagation of cascading power outages

Aidan Covarrubias, Olivia Jarrett, Christopher Paolini - San Diego State University

Cascading power outages can inflict devastating consequences on the economic security of city infrastructure and can even result in fatalities.  When demand exceeds the carrying capacity of a transmission line, that line will shut down, and a chain-reaction of propagating line outages will commence.  When a transmission line fails, the downstream network of distribution lines that deliver power to individual homes and businesses will shut down, causing loss of power to critical appliances, such as refrigeration systems, which require constant power to prevent food spoilage.  We present MIST, a fog-computing framework designed to monitor local demand and perform load-shedding decisions at the network edge. The framework consists of "smart" electrical receptacles with microcontroller measurement devices to monitor and predict the transient coverage of a transmission line outage and cooperatively shed power to halt propagation.  Shedding techniques employed by utilities do not distinguish among appliances and can result in a total power loss to selected users.  The MIST architecture will benefit Smart Cities by providing a means by which utilities can selectively shed power at the grid edge by powering off appliances in order of criticality, thereby mitigating the propagation of a failure and reducing costs that result from power loss.

Smart inverters and power electronics

A new circuit for increased fault tolerance and recovery for medium voltage multi-level three-phase inverters

Haider Mhiesan, Christopher Farnell, Roy McCann – University of Arkansas

Solar photovoltaics and wind turbine generators achieved 10% of US electricity generation in March 2017. This percentage is expected to increase in future years. The electricity generated from wind and solar sources requires power electronic circuits in order to make this energy available to the electric utility grid. Consequently, there is a need for improved fault detection and isolation in power electronic converters for maximizing the reliability of renewable energy sources. In addition, there will be power electronics needed for utility-scale battery storage systems to provide back-up energy for times of reduced wind and solar generation output and to provide black-start capability for restoration after power outages. This paper presents a new and novel power electronic circuit and method for isolating faults in power electronic inverter-rectifiers. The method is based on a modified cascaded H-bridge (CHB) topology. The modification adds one addition power device per cell.  This allows system operation to continue with of one (or more) components suffering from a faulted condition. For this research, fault types considered are either failed open or short from emitter-collector for IGBTs. Experimental results from a 2.5 kW 5-level CHB inverter are provided that confirm the benefits of the method.

Application and load specific multi-port power electronics

LIzon Maharjan, Carlos Caicedo-Narvaez – University of Texas at Dallas

Electric machines have transformed our lifestyles and as a result they are one of the most prominent electrical loads. With advances in power electronics, clean energy resources, and smart devices, the energy required to power most of the electric machines can now be sourced from a shared platform rather than using grid solely, like it has been done traditionally. This can be advantageous in several scenarios. In developed countries, the motors that are placed outdoors can be powered by solar power when enough energy is available, for example AC compressors and water pumps. This reduces the electric cost and enables utilization of clean energy resources without having to install residential level nano-grids. In developing countries, such systems can provide power to electrical loads in remote areas that do not have grid or have reliability issues. The presented study includes development of Multi-port Power Electronics devices of different capacity targeted towards following applications: i. 200W rice grinder in remote location (solar + battery), ii. 1.5hp outdoor water pump (solar + battery + grid), iii. Smart inverter and power sharing for AC compressor (solar + battery + grid + smart IOT sensors).

Design considerations of a flying capacitor multilevel flyback converter for DC-DC and pulsed power applications

Santino Graziani, Dr. Brandon Grainger, Dr. Ansel Barchowsky – University of Pittsburgh

This paper presents the development of a Flying  Capacitor Multilevel Flyback Converter (FCMFC) that provides  easily scalable peak output voltage for pulsed power. Based on a  combination of the flyback and flying capacitor multilevel  (FCML) boost converter topologies, the FCMFC achieves faster  slew rates and peak power than previous pulsed power converters.  FCMFC distributes the high voltage gain across the SDCs,  lowering device stresses and reducing the required magnetic  component size compared to existing architectures. Two  applications are addressed in this work: one for DC-DC boosting  and one for pulsed power conversion. The FCMFC achieved a  100V to 1000V step-up while only utilizing half of the electrical  hardware compared to existing topologies. Another FCMFC was  developed to produce 36kV pulses at 8kV/μs, with a 50% decrease  in the magnetic component size compared to an existing  architecture from the literature.

GaN-based point of load converters for internet of things devices

Thomas Cook1, Nicholas Franconi1, Dr. Ansel Barchowsky2, Dr. Brandon Grainger1, Dr. Robert Kerestes1, - 1) University of Pittsburgh and 2) NASA JPL

As distributed energy resources (DER) and microgrids continue to grow, the requirements levied on power-conversion electronics for system monitoring and Internet of Things (IOT) devices have become increasingly demanding. Designing for compute-intensive processing capabilities of IOT sensors further encourages the use of lightweight, compact, and efficient power-conversion electronics. When operating in a microgrid, sensors can be subject to a variety of environmental variables including radiation and temperature which can cause the power electronic modules to fail. To alleviate this disparity, this poster presents the design, development, and testing of three temperature and radiation-tolerant, point-of-load (PoL) converters using Gallium Nitride (GaN) High-Electron Mobility Transistors (HEMT) and commercial controllers to enable the success of DER and IOT devices. single phase 120V AC and produce a 3-phase output with a voltage range of 0-5V and a current range of 0-250mA. The frequency had to be able to adjust between 50Hz and 400Hz and the phase must be entirely adjustable.

Smart power electronics module for training

Chondon Roy, Namwon Kim, William Halaburda, Nathaniel Belles, Babak Parkhideh – University of North Carolina, Charlotte

This poster presents a modular power electronics education board for demonstrating the advantages of Wide-Bandgap (WBG) semiconductor devices. This instrument will act as an essential resource for training power electronics engineers and encouraging the implementation of WBG switches. The power electronics module consists of a motherboard and four plug-and-play daughter boards. This "plug-and-play" capability aids the student in comparing switching loss, conduction loss, and other characteristics between various types of semiconductor switches.