The integration of grid-forming inverter-based resources (GFM IBRs) into the transmission system is dependent on the ability of interconnection and transmission planning engineers to sufficiently represent and study their performance and impacts in both the near- and long-term time horizon.
Given that the capabilities and performance of GFM battery energy storage systems (BESS) are primarily software-based similar to grid-following (GFL) IBR which are being interconnected rapidly within the present paradigm—as opposed to other IBR technologies, like wind turbine generators, which also have hardware and mechanical considerations when delivering GFM performance—the same modeling guidance that is currently available for GFL IBR can and should also be applied to GFM IBR.
Further, given the complexities of GFM (and GFL) IBR controls and their potential effect on the grid, in order to sufficiently represent the behavior of GFM IBRs, it is increasingly critical to implement currently published guidance, such as North American Electric Reliability Corporation (NERC) dynamic modeling guidance, IBR-related alerts, lessons learned from major event analysis, and international guidance (for example: the Australian Energy Market Operator’s Dynamic Model Acceptance tests (DMAT).
The discussion below provides a general overview of IBR model types used for various study types and is applicable to both GFL and GFM models independent of simulation domain. It also includes GFM-specific considerations as well as presents the currently available GFM models.
The Need for a Comprehensive Approach to Help Ensure Reliability
The integration of GFM IBRs requires sufficiently accurate analysis of the effects of these resources on the power system in interconnection studies, local reliability[1] studies, forward-looking studies examining high-level trends and informing integration strategies on the long-term time horizon, and academic studies.
Interconnection Studies and Local Reliability Studies
Interconnection studies and local reliability studies are extremely important for the integration of IBRs, as they provide insight into the potential effects of a new resource on the power system as well as inform engineers which potential performance aspects need to be improved for the IBR to operate in a more reliable manner. In this near-term time horizon,[2] information regarding current system topology, system dispatch and resource mix, and an intended IBR plant design and original equipment manufacturer (OEM) is often known and as such, additional detail in the model space can provide significant benefits for both the interconnecting utility as well as the IBR developer.
Accurate representation of IBRs in this time horizon and during the interconnection process specifically is essential to determine what IBR performance is necessary to help improve stable operation on the power system and determine which IBR controls can be leveraged to bring about more efficient dynamic behavior and robustness. Study work at this point in the study process often results in changes in the IBR plant design and/or control parameter values that can bring about an improvement in performance at the particular site. These site-specific changes resulting from the studies should then be sufficiently mapped to site-specific changes implemented in actual equipment.
Long-Term Planning and Research Studies
Long-term planning studies and research studies are also necessary to determine high-level trends, targets, and strategies that must be implemented to prepare the power system for and integrate high levels of IBRs. In these studies, it is often not necessary or possible to model specific manufacturer-verified performance for particular products, as it is impossible to accurately estimate which manufacturers, products, interconnection points, or site-specific tuning will be used across the long-term time horizon. But studies in this time horizon can be effective without manufacturer- and topology-specific information and can provide important general insights with generic models. Since the goal for these studies is to understand general trends on a long-term time horizon, model usability and computational time efficiency can be prioritized over detailed manufacturer-specific code and high model fidelity.
Given these two very different but complementary time horizons, it is important to understand which model types are available for power system engineers to use in different types of studies as well as the strengths and weaknesses of each model type.
Differences Between User-Defined and Generic Model Types
The two model types commonly used are user-defined models (UDMs) or vendor-specific models, and standard library or generic models. Both model types are available in both the electromagnetic transient (EMT) simulation domain and the positive-sequence phasor-domain (PSPD) simulation domain—the terms describe the models themselves and not the simulation engine. Below we discuss UDM and generic models along with their strengths, weaknesses, and use cases.
Present Dynamic Modeling Recommendations and Directives for North America
It is important to keep in mind that currently published NERC modeling guidance, as well as Federal Energy Regulatory Commission (FERC) Order 2023, state that both model types should be created such that they represent the IBR and be supplied to the applicable entities by the IBR developer. This allows study engineers to use engineering judgment to use the most appropriate model type for each study.
The figure below is adapted from slides 92 and 93 from a presentation given by NERC staff at the November 19-20, 2024 joint NERC-North American Transmission Forum-Electric Power Research Institute (EPRI) Technical Workshop and summarizes some of NERC’s current modeling recommendations. While the figure references positive-sequence phasor-domain modeling, the same principles apply in the EMT domain as well. For further information, please see NERC’s dynamic modeling recommendations, Milestone 3 of NERC’s FERC Order 901 Workplan, and FERC Orders 901 and 2023.
NERC Recommendations for Dynamic Model Types
Source: NERC-North American Transmission Forum-Electric Power Research Institute (EPRI) Technical Workshop.
Standard Library (Generic) Models
Generic models are not specific to any manufacturer or, often, to a particular generating technology type. These models are created through collaboration between numerous industry stakeholders and use significant amounts of expertise, measured IBR performance data, and thorough collaboration with multiple IBR manufacturers to create publicly available and simplified models that can be parameterized to generally represent IBR technology.
Generic models have numerous benefits when used for appropriate studies and within their capabilities. They offer a standardized, modular, and well-documented model interface that is typically easier to learn and use in study work than are the manufacturer-specific nuances associated with user-defined models.
While generic models are typically easier to use, their generic nature means that they may not represent an actual IBR plant with the level of accuracy necessary to study the effects of a specific IBR plant on the transmission system. This is because they do not include any manufacturer-specific control code or logic, and they often have limitations when implementing real, on-site, control parameters into the model space and from the model space into the on-site controllers without extensive validation and analysis performed by the inverter OEM for each specific plant. Additional mapping—where possible and with significant assistance from each OEM at a given plant—is needed to implement parameter changes necessary for stability as determined in the study space into real-world IBR equipment.
User-Defined Models
User-defined models are created by each specific IBR manufacturer and are intended to be highly accurate and high-fidelity representations of the IBR technologies. These UDMs often leverage proprietary information like actual control code, proprietary capability limitations, and other intellectual property. The IBR manufacturer’s use of this proprietary information often results in models able to very accurately reflect real equipment performance under numerous tests within a test system.
If developed in a robust and careful manner, UDMs have advantages over generic models in terms of accuracy, fidelity, and the ability to map parameters between models and real equipment. These benefits are particularly valuable in interconnection and local reliability studies, as accuracy is of the utmost importance in these scenarios. Additionally, the often close mapping and high level of fidelity of UDMs allow for solutions deemed necessary for transmission system reliability to be more easily implemented in the real IBR plant.
This higher level of accuracy and fidelity does come with additional considerations that should be taken into account when choosing which model type to use for a particular study. UDMs, because they are manufacturer-specific, may have significant variability in model structure, model set-up, and operation. This can mean an additional learning curve for study engineers as they need to learn the nuances of each manufacturer and its models. The level of detail within UDM documentation packages is also widely variable, which can affect how easy it is to use a given model.
Another important aspect of UDMs to consider is the varying amount of model parameters exposed to the end user. The amount of parameters exposed to the end user drastically affects the user’s ability to alter the performance and capabilities of the modeled representation and overall model usability. The practice of exposing only some model parameters is known as “black boxing,” with the amount of black-boxing varying from manufacturer to manufacturer. UDM models may contain the full source code and all parameters that are included in the IBR, which are typically in the thousands of parameters, but not allow the end user to change many of these parameters. The vast majority of UDM parameters are product-specific and are critical to an accurate representation of limitations of the physical product (e.g., current limits, ride-through limits, etc.); therefore, changing these parameters could result in the model showing IBR performance that is physically impossible or dangerous. For this reason, OEMs often expose only parameters that are necessary for model set-up and general site-specific tuning while black-boxing parameters that should not be altered by end users.
UDMs often have limitations when interfacing with simulation software since manufacturers may not have expertise in the nuances related to efficient and robust operation within the simulation software. Further, when many UDMs are used within a study area, pollution of data read/write across models can occur if UDMs are not efficiently constructed. Also, if the simulation platform’s simulation engine techniques are not carefully considered, then direct mapping of parameter values from the real equipment to the UDM can cause errors in modeled performance.
Currently Available Generic Models for Grid-Forming Resources
In order to efficiently integrate GFM resources into the transmission system, forward-looking study work is a necessary piece to understand high-level trends and inform integration strategies. For this type of critically important study, generic models provide an easy-to-use and publicly available representation of generic GFM IBR performance and can be used to provide the industry with essential information to inform GFM IBR integration efforts. These generic GFM IBR models are becoming available to the industry at a time when GFM controls are highly proprietary, in the case of solar photovoltaic and battery energy storage, or not yet mature, in the case of wind resources, and provide the industry with the tools necessary to inform the strategies needed to implement GFM resources at scale.
REGFM_A1
The development of the REGFM_A1 model took place through a collaboration between the Pacific Northwest National Laboratory’s Directed Research and Development program and the Universal Interoperability for Grid-Forming Inverters (UNIFI) consortium. The UNIFI consortium is funded by the U.S. Department of Energy’s Office of Energy Efficiency and Renewable Energy under the Solar Energy Technologies Office Award Number 38637.
This model is a PSPD model capable of preliminarily representing droop-controlled GFM IBR characteristics and has been approved by the Western Electric Coordinating Council (WECC) Model Validation Subcommittee for use in WECC interconnection-wide studies. This model was created through collaboration between the Pacific Northwest National Laboratory, EPRI, and SMA.
The technical report Model Specification of Droop-Controlled, Grid-Forming Inverters (REGFM_A1) includes the model, parameters, and additional information.
REGFM_B1
The development of the REGFM_B1 model was supported by UNIFI, funded by the U.S. Department of Energy’s Office of Energy Efficiency and Renewable Energy under the Solar Energy Technologies Office Award Number 38637, with technical support from the WECC Model Validation Subcommittee for use in WECC interconnection-wide studies.
This model is a PSPD model capable of preliminarily representing a virtual synchronous machine (VSM) GFM inverter. This model was created through collaboration between the Pacific Northwest National Laboratory, General Electric (GE), EPRI, and Siemens Gamesa Renewable Energy and is intended to increase the industry’s understanding of VSM GFM resources.
The technical report Virtual Synchronous Machine Grid-Forming Inverter Model Specification (REGFM_B1) includes the model, parameters, and additional information.
EPRI Generic Grid-Forming Models (GNRGFM)
These models have been developed through EPRI’s participation in the UNIFI consortium and are a preliminary representation of various behaviors of the technology.
These highly configurable PSPD models are designed for representation of GFM IBRs in industry-standard software packages. These models have a few nuanced features that have been insightful in various recent system studies around the world that have aimed to identify the value proposition of GFM technology.
Additional improvements will be made to these models as EPRI learns more about the technology and receives feedback from users of the model.
The models, manuals, and example cases can be downloaded at the links below. Additionally, feel free to reach out to Deepak Ramasubramanian at dramasubramanian@epri.com regarding any feedback that you may have about this model, its applications, and its features.
PSS(R)E: https://www.epri.com/research/products/000000003002030987
PSLF: https://www.epri.com/research/products/000000003002030988
PowerFactory: https://www.epri.com/research/products/000000003002030989
The models are complementary to EPRI’s EMT domain models for GFM technology. To learn more about the EMT domain models, please reach out to Deepak Ramasubramanian at dramasubramanian@epri.com.
[1] Local reliability studies refer to any study or study area for which the highest level of accuracy should be prioritized. For example, if an independent system operator is performing a reliability assessment, NERC guidance recommends using verified manufacturer-specific user-defined models for the ISO’s footprint while allowing for the use of manufacturer-verified generic models to represent the IBR outside of the study area.
[2] Near-term time horizon refers to the interconnection process or local reliability studies (typically within five years of the study date).