The Future Power Grid
The need to include renewable energy supplies with distributed energy generation, new storage elements and the introduction of electric vehicles (EVs) demand that the future power grid will change its centralised configuration, therefore, microgrid technologies are gaining widespread interest. However much research is required to test different strategies of microgrid implementation, control algorithms, communications, power electronics, and measurement.
The University of Leeds approached Austin Consultants to develop a microgrid laboratory that provides a high-fidelity simulation platform to provide the test and measurement capabilities to investigate different microgrid configurations, scenarios such as weather phenomena, and apply technologies related to energy flux transfer, power control, communications, EVs, power systems and other technologies.
Austin Consultants were instrumental in developing the smart grid laboratory and delivering a flexible platform that ensures long term return on our investment. The laboratory facilitates research into the challenges and problems at the diverse stages of electric power systems and applications, ranging from generation and conversion to transmission, distribution, and consumption.Dr Petros Aristidou
The MicroGrid Laboratory allows users to create a custom single or three-phase electrical grid on which to run experiments and conduct research.
The lab consists of four distributed generation unit emulator (DGUE) units, provided by Cinergia, connected to a central Windows server, running a LabVIEW server application, that is responsible for all communication to and from the units, and four client PCs connected via TCP, running a LabVIEW client application, that allows the users to interact with and configure the system, set up, experiments or load in previously saved configurations.
Each DGUE is configurable through software to operate in one of three modes:
- Grid Emulation Mode (GE)
- Electronic Load Mode (EL)
- Power Amplifier/Model Mode (MM)
Model Mode allows the unit to run a simulation model (e.g. based on energy generation from a renewable source) in real time and act as a voltage or current source on the grid.
A grid requires a minimum of two DGUE units so with four available the system can be configured in two ways:
- Full grid, all 4 units in use and connected to each other.
- Two grids: 2 of the units connected to each other, and the other 2 units connected to each other.
The DGUE units each require a network connection and three phase power supply. Standard configuration requires direct physical cabling to connect the units Microgrid connections however it is possible for the end user to connect the DGUEs via their own Patch Panel or PI Models/Transmission Line Models.
The control software has been developed with research and experimental uses in mind and does not restrict the grid makeup/framework to allow the end user to customise the grid as required. Each DGUE has a range of electrical safeguards and trips/alarm to protect themselves from damage due to incompatible user configurations, such as high load.
Management & Control System
The system software framework is designed to allow future expansion of grid items, clients and additional hardware nodes to be added without requiring the core framework to be updated.
Each workstation runs a LabVIEW client application that allows the users to interact with and configure the system, set up, experiments or load in previously saved configurations. The Client Application also has a Wizard for importing Models to make this a simple process with no prior knowledge of the National Instrument Model Interface Toolkit require.
Each Client PC has two monitors and the client application is designed to make use of multiple screens.
Each of the four client PCs is connected via TCP to the system’s central Windows server, running a LabVIEW server application that allows the client PCs to communicate with the DGUE units. The server also allows the DGUE units to share data with each other which can be used in the models if required.
To allow the DGUE units to deploy a variety of MATLAB simulation models each has a National Instruments Single-Board RIO Controller (sbRIO) installed that communicates with the DSP central control board via a Serial Peripheral Interface (SPI) bus link. sbRIO runs a LabVIEW application that allows the import of MATLAB models that the DGUE is then able to deploy to the hardware in real-time. This software application is responsible for:
- Running Models on the System
- Providing the DSP device with Set Points when the Cinergia Unit is in Model Mode via SPI interface.
- Data communication to update model import and parameter values.
- Data communication to download new models to the system.
The RIO device has access to all the important signals from the DGUE in real-time, including all voltage and current measurements. Six high-speed Phase-Locked Loops (PLLs) running on the FPGA give instantaneous phase and frequency information for both voltage and current, for each phase, and this information can be used within the models. Having access to this data in the FPGA provides extra capabilities to the models, such as the ability to write power setpoints. These setpoints get converted to “in-phase” current setpoints, in real-time on the FPGA, avoiding the round-trip delay of pushing the data through a model at 1kHz.
The FPGA processes the standard inputs from the system and allows the sbRIO to deploy the LabVIEW models to the hardware in real-time. The FPGA code is easily customisable, providing more advanced features such as harmonic generation/modification, instantaneous feature, or transient capabilities.
The direct connection of the RIO device via a fast SPI link allows the user to provide voltage or current setpoints with extremely high fidelity. Model outputs (currently running at 1kHz) are interpolated into smooth requests.
Each unit produces a 1Hz data stream of all data channels, with the ability to increase the logging rate to 1KHz for selected channels. Please note, currently, there is no hardware synchronisation between the units to allow for highly synchronised 1Khz data streams, therefore a 1KHz data file is generated per unit that generates 1KHz data to log.
Log files generate a high speed and low-speed file(s) as appropriate to the channels selected and are in CSV format.
The new Smart Grid Laboratory provides modelling, analysis and control design of next-generation power systems on both the system and individual component level. This includes diverse renewable energy sources as well as applications of power electronics and electric drives.
The laboratory has been developed with research and experimental use in mind, the software does not restrict the grid makeup/framework to allow the end user to customise the grid as required.
The system features:
- An adaptable MicroGrid test platform for research and experimentation.
- Customisable System Control Software.
- Reconfigurable distributed generation unit emulator (DGUE) nodes, provided by Cinergia
- Configurable fault insertion.
- Standard power units from 7.5 to 160kW.
- Real-time MATLAB/Simulink model execution at speeds up to 1KHz
- A wide range of off-the-shelf MATLAB models developed by Cener (National Renewable Energy Centre Spain). (e.g. Photovoltaic System, Wind
- Turbine, Storage System, Electric Load, etc.) available if required.
- Templates provided for custom model development.
- Extensible, flexible architecture to allow for future development, by simply adding more distributed generation unit emulator (DGUE) units.