You are here

Grid-Connected Advanced Power Electronic Systems (GRAPES)

University of Arkansas Fayetteville

University of South Carolina

University of Wisconsin Milwaukee

Last Reviewed: 06/15/2017

Research focuses on design, development, evaluation, control, and standardization of grid-connected power electronic equipment on both the supply and load side of power systems.

Center Mission and Rationale

The mission of this center is to accelerate the adoption and insertion of power electronics into the electric grid in order to improve system stability, flexibility, robustness, and economy. We expect to accomplish that mission by focusing on the following main objectives:

  • To develop new technologies for advanced power electronic systems in the areas supporting grid connected distributed energy resources, power steering and routing devices, and intelligent load-side devices,
  • To develop the software and tools for controlling embedded- and grid-connected power electronics to benefit the grid as well as controlled loads,
  • To educate engineers who understand the power electronic technologies important to the member companies.

The electric power industry is ultra-critical to the economy and security of the United States. Without electric power, everything stops - literally everything - and these interruptions have severe economic consequences. The demand for electrical energy is increasing and political and environmental pressures are forcing adoption of new distributed generation resources, such as wind, solar, and tidal, that do not fit well into the traditional architecture of the electric power grid. Robustness of the national power infrastructure is threatened by aging equipment, lack of integration between generation, transmission, distribution, and utilization, and by terrorism. Paradoxically, one of the direct near-term threats is imposed by increases in local, environmentally-green generation technologies that can undermine the traditional safety mechanisms. This threat occurs partly because the mathematical and physical structure of the power grid is evolving from a paradigm of "a few controls on systems governed by the laws of physics" to "many independent controls on systems being governed by independent digital controls". Our ability to predict the behavior of the system, and thus to control it, will rapidly erode without greater standardization of the control mechanisms embedded in power electronics. We currently have a narrow window of opportunity to re-define the power grid to improve its robustness. This is a focus of our research and is precisely where university research and industrial partnering will pay huge dividends.

Research program

Distributed Power Quality Improvement using Power Electronics and Digital Signal Processing (GR-16-03)

This project addresses power quality improvement for compensation of non-periodic load currents using sharing among distributed power electronic converters. A new technique is under development for load power quality improvement using three co-located power quality conditioners. Using simulation based on real-world data, compensator control methods have been developed for compensation and power quality improvement of highly distorting loads, such as those found in steel mills. The compensator consists of three co-located devices with different calculation windows, called fast compensator, reactive compensator, and slow compensator. Each one of them is responsible for the compensation of one phenomenon in the non-periodic current: sharp edges, reactive current, and low frequency modulation. In order to improve the flexibility of the technique, a fuzzy based adaptive window is used for the slow compensator to find the optimum window for different load characteristics. In the current stage of this project, a prototype demonstrator is under construction for experimental validation of the proposed method.

Future Hybrid Microgrids (GR-14-08)

Microgrids enable integration of distributed energy resources and increase the reliability of the power grid. Power electronic interfaces are the key components of future microgrids since power flows are subject to ac/dc, dc/dc and dc/ac conversions.

This project includes the design and construction of hybrid-microgrid prototypes based on power electronics in the MVA power range which have efficiency and system complexity issues not encountered in low-scale prototypes. As the power ratings of power electronic interfaces increase, their base impedances decrease. Thus, ac microgrid stability becomes compromised by the low-pass filter resonant propagation. To this effect, a scaled-down microgrid prototype has been built for verification purposes and safely testing the proposed ac/dc converter control algorithms. The high-power equipment (i.e., converters, transformers, circuit breakers, etc.) at the National Center for Reliable Electric Power Transmission have been modified to form a microgrid testbed. A new microgrid infrastructure is under construction for proving the new control algorithms, which aim at mitigating the high-power microgrid stability issues.

 

PMU Role in Evaluating PV Generation Impact on Transmission Grid (GR-15-05)

The increasing adoption of MW utility scale solar photovoltaic (PV) arrays presents challenges to existing electrical distribution systems. Large scale solar PV arrays may be located in areas where the feeder design was based on unidirectional power flows. With distributed PV generation, there may be disruptions to systems protection and compensation equipment. This project investigates the use of distribution-level phasor measurement units to monitor and control distribution systems that include large PV sources in order to develop methods of mitigating voltage disruptions. The result is an understanding and recommendation for the use of real-time PMU information to control the local distribution and transmission system using FACTS and D-FACTS equipment to compensate for the effects of solar PV generation.

Physics-based Analytical and Compact Modeling of GaN Power Devices for Advanced Power Electronics (NSF-15-06)

Advances in wide bandgap materials such as SiC and GaN have led to substantial advances in power semiconductor devices and are now positioned to dominate the next generation of power electronics replacing silicon devices. This research focuses on the creation and validation of analytical models for state-of-the-art GaN power devices. The market share of GaN power devices is expected to reach a staggering $15.6 billion by 2022, mainly due to the growing demands in the power and energy sector, the communication infrastructure sector, and the power electronics market. GaN devices are expected to reduce overall energy conversion losses down to 1%, resulting in annual savings of nearly $40 billion in US revenues. A high-efficiency and green energy infrastructure is vital for reducing overall expenditures and reducing the carbon footprint of the electronics industry and the environment. The expected outcome from this fundamental research focuses on developing physics-based compact device models for circuit simulations that will help electronics engineers rapidly develop circuit designs and prototypes based on GaN devices. Impacts of this model will enable a side-by-side comparison of GaN and silicon devices at the design and analysis phase. This, in turn, will likely promote increased usage of GaN semiconductor technology. The models generated in this research will be open access and made publicly accessible on the NSF Industry/University Co-Operative Research Center website under the Grid-Connected Advanced Power Electronics Systems (GRAPES) center site.

Conventionally, the GaN device is a normally-on device. GaN devices for power electronics applications are modified with a p-type GaN gate and an AlGaN buffer layer. The discontinuity in polarization between the AlGaN barrier layer and p-GaN cap layer brings about the desired normally-off operation by lifting the conduction band above the Fermi-level.

Mobile Power Substations (GR-15-03)

Developing power grids that are resilient under disruptive events is one of the main objectives of electric utilities. A light-weight mobile power substation connecting two distribution feeders having different voltage levels would be a useful piece equipment to be deployed under emergency conditions. To this end, the main goal of this project is to evaluate potential designs for a mobile power substation characterized by its light weight so it can be transported in a single truck to interface two medium-voltage distribution systems operating under emergency conditions. The research team would evaluate arrangements providing electric isolation or not. Electric isolation will be implemented through the use of a medium-frequency transformer. Initial research will be centered on the modular multilevel converter (MMC) used in HVdc terminals since it may lead to a design with the highest power density.

Multi-Port Bi-Directional Resonant Solid State Transformer (GR-16-06)

There is a shift in the decades-old paradigm of energy generation and distribution. The emerging concept includes new elements such as Distributed Generations (DG), energy storage, DC systems, and power electronics-based systems. The conventional 60Hz transformers cannot meet the flexibility and controllability demanded by this new paradigm. The goal of this project is to develop the concept for an efficient medium voltage Solid State Transformer (SST) to enable smart and reliable Distribution Systems (DS) for grid power. Many researchers have worked on the SST concept. However, this enabling technology did not make a significant penetration into the utility grid due to several drawbacks: low efficiency, low voltage/power capabilities, cost and resilient packaging of enabling high band gap devices, and concerns regarding fault protection.

The objective of this project is to use Medium Voltage (MV) Wide Band Gap (WBG) devices (i.e. SiC switches and diodes) to increase both SST voltage and power so it can be applied at the DS level. In addition, a novel resonance feature has been added to the SST concept to significantly increase its conversion efficiency. This resonant operation along with proper controls enables usage of low cost no-load disconnects or breakers for system fault protection. Detailed analyses of the proposed system will be performed both for the power electronics system design as well as application and integration in a DS.

There is an impending need in future electrical distribution systems for flexible, controllable, compact, and efficient medium voltage transformers. There are many opportunities for commercialization for the proposed system if the efficiency can be increased to 98%-99%. Potential applications include all of the utility distribution systems, microgrids, AC/DC networks, DC data centers, etc.

Distributed Energy Resources: A Testbed for Distributed Autonomous Control Concepts for High-Power Microgrids (GR-17-10)

Both the University of Wisconsin-Milwaukee (UWM) and the University of Arkansas (UA) have worked on several microgrid controls projects including high-power microgrids, hierarchical control, virtual droop control, and central control.  There are several research tasks within Project GR-17-10 to be performed jointly by UWM and UA, (i) to develop the concept for distributed microgrid controls, (ii) to evaluate the reliability improvement using distributed controls, (iii) to build an HIL setup to test and implement microgrid control, (iv) to implement a high-power microgrid  testbed (MGTB) at the UA National Center for Reliable Electric Power Transmission (NCREPT), and (v) to develop autonomous and predictive concept in a microgrid with higher penetration of renewables.  Tasks I, ii, and v will be performed at UWM, led by Prof. Adel Nasiri.  The concept of the distributed control system is based on installing fast and low cost controllers at each distributed source or smart load.  The reliability assessment will be conducted using Markov Chain theory.  Both UWM and UA will perform task iii on different platforms, with UWM on NI CompactRIO-based system and UA on Typhoon-based system.  UA will perform task iv using the existing three back-to-back voltage-source converters, the so-called regen benches that will be connected in parallel to the point of common coupling in order to emulate different types of generators and loads.  These regen benches would emulate wind power, photovoltaics arrays and other generators to determine their interaction and stability problems in high-power microgrids.  The regen benches would work in two modes: the grid-connected mode and island mode.  The UWM controller will be implemented on a system with real renewable sources and loads.  The controller will take into account the forecast for renewable energy generation and load to minimize the stress on energy storage and improve power quality in the microgrid.  The ultimate goal of this project is to compare the performances and differences between the UA high power testbed and the UWM testbed with high renewable penetration so a set of guidelines could be produced.

SiC-Based Direct Power Electronics Interface for Battery Energy Storage System into Medium Voltage Distribution System (13.8 kV) (GR-17-03)

This project involves the design and construction of a SiC-based direct power electronics interface for a battery energy storage system (BESS), which is to be integrated into a 13.8 kV medium-voltage distribution system.  Normally, to interface a BESS to a medium-voltage distribution line, a step-up transformer is required to boost the inverter ouput voltage.  The use of the transformer provides convenient isolation, however using a transformer to meet medium-voltage inverter insulation requirements leads to substantially higher leakage inductance, increased switching losses and limited transformer power transfer capability.  Recent advances in high voltage power semiconductor devices, medium-voltage (>10 kV) SiC power modules present an opportunity to realize a transformerless interface.  Transformerless topologies, and the use of wide bandgap devices, have the potential for reducing cost and size of passive components for the medium-voltage inverter.  To satisfy the medium voltage basic insulation level (BIL) requirements for the power electronics interface, modular multilevel cascade (MMC) inverters provide a better solution.  This battery energy storage system interface will also include fault protection circuitry and communication protocols.  The performance of the control algorithms for a BESS equipment will be tested through an experimental prototype at the National Center for Reliable Electric Power Transmission (NCREPT) using the 13.8 kV distribution system.

Coordinated Optimal Voltage Regulation for the Next-Generation Distribution Grids with High Penetration of PV Generation (GR-17-08)

Photovoltaic (PV) generation has been extensively deployed in the modern distribution systems.  However, high penetration of PV generation also brings about severe challenges to the grid operations.  Among all the challenges, voltage violation is the most critical, since the current voltage regulation schemes are designed to manage one-way power flow and cannot easily accommodate the fast changing dynamics in the distribution grids.  In addition, the existing grid infrastructures are ill-equipped to gain real-time visibility of distributed PV generations, since the data acquisition and monitoring systems typically do not extend beyond substations and/or distribution feeders and are not designed to handle real-time processing of large volumes of data.  To address these issues, a coordinated optimal voltage regulation (COVoR) framework is proposed to enable high penetration of PV generations.  To accomplish this goal, three specific objectives are expected to be achieved.  Firstly, we envision a self-sensing network enabled by the sensing and communication capabilities of smart inverters.  Based on these measurements, a scalable and optimal scheme will be developed to partition the distribution grid into dynamic voltage regulation (VR) zones.  Secondly, we will develop an advanced multi-agent system based cooperative control method for reactive power sharing among PV inverts within a local VR zone.  Thirdly, we will fully exploit and upswing the advanced grid supportive capabilities of smart inverters by using model predictive control.

High Step-Up/Down Transformerless Modular-Multilevel DC-DC Converter (GR-16-02)

This project develops and builds a high step-up/down transformerless dc-dc modular multilevel converter (MMC) that would be applicable to MV distribution-level power systems. The design achieves high voltage ratios for interfacing renewable energy sources such as photovoltaic, wind turbine and line interactive UPS systems. The converter uses an MMC approach operating in resonant mode in order to improve overall efficiency. This topology operates to step-up the input voltage with 1:10 or larger conversion ratio. As a bidirectional converter, it also provides step-down capability at the same voltage ratio (10:1 or greater). By eliminating the presence of a magnetic core transformer as used in conventional designs, this project provides a small, low-cost, direct, and simple solution for high step-up/down converters while meeting the safety and isolation requirements given by IEC and UL standards.

A High Power Real-time Photovoltaic Source Simulator (GR-16-04)

High power, e.g., 1 MW, photovoltaic (PV) source simulators can be utilized to evaluate the performance and study the grid integration issues of the utility scale PV inverters in the laboratories. However, due to the high cost of commercial PV simulators at MW level, which are usually programmable DC power supplies, this testing capability is not common in public testing facilities. In this project, a hybrid PV simulator is proposed to emulate PV arrays up to MW scale. The reference curves can be either generated by using an actual PV cell to ensure the high fidelity, or obtained by using model based methods, such that repeatable results can be produced. The power stage will consist of a grid-connected active front end and an interleaved dc-dc converter. A novel sliding mode controller will be developed to ensure the reference tracking performance and the bandwidth of the PV simulator. Both hardware-in-the-loop simulation and experimental studies will be performed to validate the effectiveness of the proposed PV simulator.

Fault Protection and Coordination in a DC Community Microgrid (GR-16-05)

Community microgrids have emerged as an alternative to address the rising societal demands for electric infrastructures that are able to provide premium reliability and power quality levels while at the same time being economically and environmentally friendly.  The focus of this project is a community microgrid that supplies electricity to a group of houses within a neighborhood or several connected neighborhoods in close proximity. Such a system provides a unique opportunity for everyday consumers to take advantage of renewable energy resources, such as solar through shared use.  Further benefits come through inter-connection of DC enabled smart homes which have the best chance of driving towards net zero energy usage.  The benefit of DC interconnection of homes through a microgrid is lower cost and less complex integration of multiple shared energy sources and integration with energy storage.  However, the most significant roadblock to such systems is the availability of safe and reliable protective distribution equipment. In conventional AC distribution, fault current is limited by the source impedance of the upstream distribution feed and the closer a fault is to that feed, the higher the fault current will be.  Radial distribution of circuit breaker protected branches from the transformer feed to a house and then to the individual loads provides is a time-proven method for isolating a fault closest to its location.  A DC fed home has very different characteristics when fault behavior is considered, especially if the DC distribution includes multiple sources of power such as Solar PV, Battery back-up, and DC converted utility feed. If a near zero-ohm fault is suddenly applied, the fault characteristic is dominated by energy storage on the bus and inter-connecting cables. So effective DC protective circuits must be able to discern faults and isolate them from the rest of the system on the order of microseconds. The purpose of this project is to develop and test solid state circuit breaker based radial distribution systems that can act to isolate faults with minimal need for sensing circuitry and without inter-device communications.  A unique approach is proposed which utilizes normally-on Wide Band Gap (WBG) Silicon Carbide (SiC) JFET or Gallium Nitride (GaN) HEMT devices as the fault interrupting solid state switch and a fast-starting isolated DC/DC converter as the protection driver.  

The new SSCB detects short circuit faults by sensing its drain-source voltage rise and draws power from the fault condition to turn and hold off the SiC JFET. This new circuit breaker technology offers a reaction time of 1-2μs, about 10X faster than any previously reported solid state circuit breakers and 10,000X faster than any mechanical circuit breakers.

Optimized Gate Drivers for High Voltage Power Devices (GR-17-04)

This project’s main focus is to develop a gate driver with an integrated power supply to drive high-voltage silicon carbide (SiC) devices.  In particular, the focus is on the 10 kV SiC MOSFET, which is available and has been tested in some literature studies.  The capability of commercially available gate drivers does not meet the requirements needed to efficiently drive SiC devices at the 10 kV voltage level.  However, the use of high-voltage SiC devices in power electronics is increasing.  This calls for the development of research techniques and growth in the area.  Thus, this project aims to develop and optimize a gate driver board for the 10 kV SiC MOSFET with the goal of optimizing the performance, cost, and size.

The scope of the project addresses the main issues inhibiting the development of the SiC device gate drivers, such as isolation, dv/dt EMI tolerance, and protection.  The small collection of research which analyzes the performance and characterizes the high-voltage SiC MOSFET is used to determine the gate driver’s needs.  Ongoing research and industry needs are considered in the optimization of the gate driver board design.  A PCB will be fabricated for the design, and the testbed for the module will be created.  The design cycle will consist of both simulated and physical testing, including the development of a double-pulse test at high-voltage to be done at the National Center for Reliable Electric Power Transmission (NCREPT).  This testbed development will also serve as a standard for future research projects in this area.  In addition to the optimization of the main driver functions, alternate laminate technologies will be considered including the use of LTCC (low-temperature co-fired ceramic), which would increase voltage isolation.

PV Inverter Control to Sustain High Quality of Service (GR-14-05)

Distribution systems are being challenged by voltage fluctuation due to the increasing penetration of distributed photovoltaic (PV) generation. The overall goal of this project is to define strategies for the planning, control, and coordination of PV plants that take into consideration quality-of-service requirements. We have been conducting research on the following topics:

  • A stochastic approach to optimum placement of photovoltaic generation in distribution feeders. We apply a stochastic method based on kernel density estimation to identify the optimum location of PV generations in distribution feeders. Probabilistic load flow is used to take into consideration the uncertainty of the PV output and the load demand. A 38-node distribution system located at Walterboro, SC is modeled to test the proposed algorithm.
  • Day-ahead optimal scheduling of PV inverters, OLTC and capacitor banks in distribution feeders: A day-ahead optimal scheduling for reactive power of PV inverters and tap position of on-load tap changers (OLTC) is proposed. The objective is to minimize the node voltage deviations and power losses while keeping the frequency of the intraday tap changer operations under a predefined value. The PV power and load demand are forecasted day-ahead. Simulations are conducted of a modified IEEE 34 node and IEEE 123 node distribution system with and without the optimal control. A Monte Carlo-based probabilistic load flow simulation is proposed to compare the results with and without forecast errors on the PV generation output and load demand.
  • Real time control of PV inverters. To compensate for hardly predictable generation and load variability we are developing a fast real time control that corrects the setting of the day- ahead optimal scheduling. We are testing this system using our laboratory infrastructure in a hardware in the loop approach.

 

Fault Detection and Management Needs Development Protective Relaying Methods for Microgrids (GR-17-02)

Because the microgrid is a dynamically changing mesh that will respond differently to faults depending on its configuration, achievement of reliable fault discrimination drives complexity and cost.  When short circuit faults occur within a microgrid multiple sources of energy can feed the fault, including adjacent electronic loads with front-end filter/storage capacitors—this is particularly the case with DC microgrids where sudden fault inception is characterized only by connected capacitors and cable inductances.  An array of additional corner case scenarios exists, each of which must be handled in a different way.

This project is a collaborative effort between UWM (Cuzner) and USC (Ginn, Benigni) to develop Hardware in the Loop (HiL) and Power Hardware in the Loop (PHiL) test platforms to develop protective relaying approaches for AC, DC and hybrid AC/DC microgrids.  Presently, UWM has developed a Controller-Hardware in the Loop (CHiL) system that enables the study of timing propagation delays between distributed controllers embedded within Distributed Energy Resources (DERs), reliability of a decentralized microgrid control architecture and demonstration of scalability concepts.  The UWM CHiL consists of Compact RIO units used to collect feedback information and interface with a Tertiary controller implemented in LabView.  USC has developed an Integrated Grids Laboratory (InteGraL) that supports combined simulation of power and communication grids for testing distributed solutions for control and monitoring in distribution grids.  The InteGraL system uses OPAL RT for power system simulation, NS3-RT and Apposite N-91 for communication network emulation, Compact RIO to emulate distributed control and data collection interfaces and multi-purpose ARM based processors to augment the HiL real-time simulation capability, 10 Gbith communication between nodes is available.  The plan is to augment the CHiL and PHiL systems at UWM and USC to add high speed serial communications for protective relaying. 

Common FPGA-based high speed serial communications implementations will be incorporated into the communication network emulations in order to research distributed and centralized schemes for achieving fault discrimination within the microgrid and to develop self-healing systems having autonomous fault detection, isolation and reconfiguration capabilities.  This effort is part of a wider vision to enable collaborative research encompassing all levels of microgrid systems to control an application to various industries.

Special Activities

Associated personnel have extensive research programs in: 

  • Power systems, including off-grid systems such as in transportation systems (ships, planes, trains, and automobiles)
  • Power electronics devices, characterization, modeling
  • Control systems and motor drives
  • Integrated and power electronic circuit design and testing
  • Simulation  methods and environments for multidisciplinary dynamic systems

Facilities and Laboratory

 

National Center for Reliable Electric Power Transmission, University of Arkansas, Fayetteville

  • $4.5 M facility opened in 2008
  • 7,000 sq. ft. with four test bays
  • Tractor trailer accommodation for large equipment under test
  • Variable Voltage to 15 kV
  • Variable Frequency
  • Reconfigurable
  • Programmable
  • Uniquely Capable

   

Virtual Test Bed Facility, University of South Carolina, Columbia

  • Power Routing
  • Power Electronics
  • Realtime Electro Mechanics
  • Virtual Test Bed Development

   

University of Wisconsin, Milwaukee Facilities

  • 350kVA, 480V, Microgrid facility
  • 50kW solar PV
  • 12kW wind turbine
  • Two 45kVA generators
  • 90kWh, 125kW energy storage
  • 50kWh, 25kW flow battery
  • Active and passive loads
  • Reconfigurable 
  • Flexible control
  • Energy storage test setups

Locations

University of Arkansas Fayetteville

3217 Bell Engineering
1 University of Arkansas

Fayetteville, Arkansas, 72701

United States

(479) 575 - 4838

(479) 575 - 7967

http://grapes.uark.edu

University of South Carolina

Swearingen Engineering Center
301 South Main Street Room 3A80

Columbia, South Carolina, 29208

United States

University of Wisconsin Milwaukee

3200 N. Cramer Street
Rm. E371D

Milwaukee, Wisconsin, 53211

United States