Power Systems Simulation
Power Systems Simulation
The Power Systems Simulation team focuses on the development of standardized simulation model exchange in order to couple third-party simulators through a common interface.
The Role of Simulations
Simulations are one of the most important tools for researchers, engineers, and other stakeholders to assess, design or evaluate new technologies. Those technologies can be new resources, e.g., different types of energy storage or electricity generators, or novel control systems, e.g., to increase efficiency or add additional functionality.
Simulations also allow the exploration of different scenarios, e.g., high penetration of renewable energy resources with and without co-located storage, to evaluate the interdependency of coupled systems in the context of local and global coupling of electricity grids. However, most industry standard simulation tools are domain-specific and are specialized for certain applications. For example a distribution modeling tool, as typically used by U.S. utility companies, provides good capability for distribution system planning and evaluation; but may lack the capability of emerging technologies, e.g., predictive control algorithms.
What We Do
Berkeley Lab’s Grid Integration Group (GIG) is adopting the Functional Mock-up Interface (FMI) (fmi-standard.org/) as a free standard that defines an interface to exchange dynamic models using a combination of C code and XML files. It originated in the European automotive industry and is currently supported by 100+ tools. Berkeley Lab has developed a toolchain to export, co-simulate, and analyze third-party simulators (see tools and functionality below).
Projects
SimulatorToFMU (github.com/LBNL-ETA/SimulatorToFMU) is a tool to exploit the Python Application Programming Interface (API) of third-party simulators to wrap them in compliance with FMI.
Smart Control of Distributed Energy Resources (SCooDER) (github.com/LBNL-ETA/SCooDER) is an open-source library of various simulation models relevant to power systems. It includes detailed and simplified models of advanced grid sensors, smart inverters, transformers, Photovoltaic (PV), Battery Energy Storage (BES), Electric Vehicle (EV), voltage dependent controllers, and grid models. While some models are modified versions of existing models from the Modelica Buildings Library, most models were newly developed, and validated using LBNL’s FLEXLAB/FLEXGRID (flexlab.lbl.gov/introducing-flexgrid) testbed.
Cyber Physical Co-simulation Platform for Distributed Energy Resources (CyDER) (github.com/LBNL-ETA/CyDER) is an open-source simulation platform to couple any number of (co-)simulators, and coordinate their execution. It was developed by Berkeley Lab and used in a variety of projects which scale from a single site installation with PV and BES, up to high-fidelity simulation of a U.S. state’s electricity grid, representing about 4 million individual customers’ PV system and smart inverter.
Additional developments include the pmu_event_library (github.com/LBNL-ETA/pmu_event_library) which holds recorded data samples of grid events, the Synthetic Ancillary Service Generator (SynAS) (github.com/LBNL-ETA/SynAS) which can generate frequency regulation signals, and the Extremum Seeking Library (github.com/LBNL-ETA/ESL) which includes adaptive control systems.
SimulatorToFMU (github.com/LBNL-ETA/SimulatorToFMU) is a tool to exploit the Python Application Programming Interface (API) of third-party simulators to wrap them in compliance with FMI.
Cyber Physical Co-simulation Platform for Distributed Energy Resources (CyDER) (github.com/LBNL-ETA/CyDER) is an open-source simulation platform to couple any number of (co-)simulators, and coordinate their execution. It was developed by Berkeley Lab and used in a variety of projects which scale from a single site installation with PV and BES, up to high-fidelity simulation of a U.S. state’s electricity grid, representing about 4 million individual customers’ PV system and smart inverter.
The pmu_event_library (github.com/LBNL-ETA/pmu_event_library) which holds recorded data samples of grid events.
The coupling of simulators, as shown in Figure 1, is implemented as a socket communication where the simulator, in this case Pandapower, is wrapped in a Python wrapper with a standardized exchange function. The SimulatorToFMU tool is used to automatically setup the socket communication and expose the standardized C functions to interface with FMI.
Figure 1: Export of third-party simulator using SimulatorToFMU
A model which was exported in compliance with FMI is called a Functional Mockup Unit (FMU). Depending on the level of support, simulators may provide a direct export of FMU, a Python API which can be exploited with SimulatorToFMU, or customized C functions. CyDER can then plug-in any number of FMUs, as shown in Figure 2, and automatically link them through a connection list. It coordinates the execution of FMUs and provides the final result for the coupled system.
Figure 2: Overview of the CyDER platform with FMI/FMU interface
While CyDER was developed for single-workstation application, additional developments enabled the parallelization of FMUs which allowed the usage of Lawrencium (sites.google.com/a/lbl.gov/hpc/) High Performance Computing Cluster (HPCC). Berkeley Lab was able to demonstrate a scaleup to a U.S. state’s level using a hierarchical parallelization approach, shown in Figure 3.
Figure 3: Hierarchical parallelization framework for simulation at HPCC
Meeting California’s statutory energy goals requires significant increases in the number of plug-in-electric vehicles (PEVs) and renewable power generation. PEVs can provide energy storage needed to support renewable power generation through vehicle-to-grid (V2G) and vehicle-to-building (V2B) services. The risk of accelerated electric vehicle battery degradation is commonly cited as a concern inhibiting the implementation of V2G and V2B technology. However, little quantitative evidence exists to refute or substantiate these concerns for different grid services. Repurposing PEV batteries at the end of their useful life for transportation, commonly referred to as “second-life”, can have significant impacts on reducing greenhouse gas (GHG) emissions. This project will develop optimization and control methods for providing V2G and V2B services while minimizing PEV and second-life battery degradation. The poster will present the design of a controlled study of the impact of providing V2G/V2B services on PEV batteries that uses new batteries in an aggregated fleet of Nissan LEAFs currently operating at the Los Angeles Air Force Base (LAAFB). The project will in parallel develop a scalable second-life battery energy storage solution for fleet applications, and demonstrate the integration and application of second-life battery storage into V2G/V2B services, including load shifting and those that improve power quality for customers with on-site photovoltaic (PV) generation.
Smart Control of Distributed Energy Resources (SCooDER) (github.com/LBNL-ETA/SCooDER) is an open-source library of various simulation models relevant to power systems. It includes detailed and simplified models of advanced grid sensors, smart inverters, transformers, Photovoltaic (PV), Battery Energy Storage (BES), Electric Vehicle (EV), voltage dependent controllers, and grid models. While some models are modified versions of existing models from the Modelica Buildings Library, most models were newly developed, and validated using LBNL’s FLEXLAB/FLEXGRID (flexlab.lbl.gov/introducing-flexgrid) testbed.
The Synthetic Ancillary Service Generator (SynAS) (github.com/LBNL-ETA/SynAS) which can generate frequency regulation signals.
The state of Indiana has determined that understanding the impacts that upcoming fuel transitions and emerging technologies will have on their electrical grid is critical to maintaining a low-cost, high reliability power grid. To evaluate the anticipated impacts, Indiana commissioned LBNL and Nexant, Inc. to perform a technical, economic, and reliability impact study on emergent distributed energy resources (DERs), focused on their potential impacts by 2025 and 2040.
LBNL performed simulations predicting the impact of rooftop solar systems (PV), electric vehicle charging (EV), and battery storage systems installed in both residential and commercial buildings. The evaluations determined 1) The likely physical impacts on distribution, transmission, and generation capacity requirements, 2) The impacts on utility rates for customers and grid-wide economic impacts, and 3) The reliability impacts on the system as a whole. LBNL evaluated six different adoption scenarios combining different deployment levels of the three studied DERs and other grid effects.
The study determined that the economic impact of increased DER adoption within Indiana’s IOU service territories will range between -$265 million and +$105 million in 2025, and between -$550 million and +$1.6 billion in 2040. The resiliency simulations evaluated the impact of behind-the-meter battery installations on mitigating power interruptions. Results showed that battery adoption will have limited effects on grid outages.
For more information see the full technical report: Carvallo, J., Collins, M., Bieler, S., Mueller, J., Gehbauer, C., Larsen, P. (2020). Indiana 21st Century Energy Policy: Emerging Technologies on the Electricity Distribution System. Impact on Rates, Reliability, and Resilience. Lawrence Berkeley National Laboratory.
and derived publication highlighting the findings here: Carvallo, J., Bieler, S., Collins, M., Mueller, J., Gehbauer, C., Gotham, D., Larsen, P. (2021). A framework to measure the technical, economic, and rate impacts of distributed solar, electric vehicles, and storage. Applied Energy.
Distributed Energy Resources (DERs) cause power to flow from the distribution grid up to the transmission grid, causing voltage violations. The U.S. Department of Energy Solar Technologies Office commissioned UC Berkeley's California Institute for Energy and Environment (CIEE) and Lawrence Berkeley National Laboratory (LBNL) to develop a novel controller capable of mitigating these issues. This project developed a phasor-based controller (PBC) that utilizes various DERs throughout the utility grid to control the magnitude and phase (known as the “voltage phasor”) of a chosen bus. The phasor at that bus is measured with a 𝜇PMU and compared to the phasor at a reference node (typically the substation node). The PBC utilizes a two level controller to optimize system operation. The two levels of the controller are:
- Supervisory layer: The supervisory layer (SPBC) coordinates the control of the grid from a high-level, identifying the optimal bus-level solution and sending target phasors to lower level controllers.
- Lower level controllers: The lower level controllers (LPBC) receive the targets from the supervisory layer and compute actuation signals for each device they control.
LBNL provided expertise in thorough testing of the PBC using the FLEXGRID Power Hardware-in-the-Loop facility. The tests studied the performance of the controller in systems with 1) Three photovoltaic solar (PV) systems, 2) Three Lithium ion (Li-Ion) stationary batteries, and 3) a real-time simulator coupled to a grid emulator.
Experiments utilized both Controller-In-The-Loop and Hardware-In-The-Loop configurations to test the PBC in a 33 node model, an unbalanced 13 node model, a balanced 13 node model, and when connected to a real distribution feeder. In all cases the PBC demonstrated an ability to successfully track the phasor target, validating the novel control approach.
For more information see the published paper: Baudette, M., Swartz, J., Moffat, K., Pakshong, J., Chu, L., Gehbauer, C., von Meier, A. (2020). Hardware-In-The-Loop Benchmarking Setup for Phasor Based Control Validation. Presented at Modeling, Estimation, and Control Conference October 2021
