Thrust 2: Model and integrate inverter-based resources

The Goal

To inform high-fidelity 100% renewable energy grid (T100RE) models with dynamic modeling, theory, and associated tools for stability analysis and control of T100RE power grids to leverage inverter based resources (IBR) flexibility for ancillary services.

The Theory

A significant fraction of new clean energy resources will be interfaced with the power grid via inverters whose dynamic response is dictated by vendor-specific software control rather than the physical laws governing the behavior of the resource itself. While fast controllability of IBRs presents an opportunity to shape the power grid dynamic response, there is a risk of triggering unstable or poorly damped oscillations and unacceptable variations in grid frequency. Such stability threats force power grid operators in Australia, the UK and the US to operate conservatively by curtailing variable renewable energy (VRE) outputs. Furthermore, as IBRs displace fossil-fuel generators, the ancillary services provided by the latter to maintain power grid stability diminishes. Without enabling the provision of ancillary services from IBRs, fully decommissioning fossil-fuel generators is impossible, or would be prohibitively costly.

The Methods

Section 1: Dynamic modeling of IBR-dominated power grids

We will develop models and parameter estimation algorithms that can accurately capture IBR-driven instabilities and trace their root cause for effective mitigation. At present, high-fidelity electromagnetic transient (EMT) simulations are arguably the only way to capture such instabilities, which are computationally expensive and difficult to scale and cannot readily identify the root cause. In principle, phasor/Root-Mean-Square (RMS) models of IBRs should capture such IBR-driven instabilities with much less computational burden, if the model structure is adequate and the model parameters reflect operating-point dependency. With Project Affiliates National Renewable Energy Lab (NREL) and Denmark Technical University (DTU), we will seek to identify phasor/RMS models with a sufficient level of dynamic complexity that allows them to (a) accurately represent power grid behavior and (b) are computationally tractable using singular perturbation method. We will also implement data-driven system identification methods to track, in real-time, the dynamic model of an IBR-dominated power grid. The IBR models will be validated to track dynamic conditions in the T100RE power grid, capture sub-synchronous instabilities and locate its root cause allowing for effective mitigation.

Section 2: Stability Analysis Under Uncertainty

We will develop a small-signal stability analysis methodology that can account for the modeling uncertainty induced by IBR’s vendor heterogeneity and variable operating conditions in a T100RE power grid. This work will leverage recent foundational results in network control that enable the design of scale-free analysis tools to guarantee power grid stability in a plug-and-play setting, which is particularly important for the large-scale integration of DERs. Specifically, these results ensure the stability of interconnected resources for models that are only accurate under “strong” grid assumptions. We will extend the results of for accurate power grid models under a wide range of grid strength. As part of this task, we will use the models developed in RD 2.1 to establish conditions on the input-output response of generic IBRs that (a) depend on the network conditions and (b) can guarantee stable interconnection of IBRs and synchronous generators. We will extend the analysis to provide efficient metrics of robustness that can measure stability margins, i.e., to assess and tradeoff between the level of uncertainty in different parts of the grid with the level of conservativeness in parameter tuning. This RD will yield decentralized stability analysis algorithms, based on input-output properties of individual IBRs and other grid components, that assess operating-point-dependent stability and can act as early warning for T100RE power grid operators in the control room,

Section 3: IBR Control Design for Voltage and Frequency Shaping

We will redesign IBR control loops, for both grid-following (GFL) and grid-forming modes (GFM), to better leverage their flexibility, when available, for mitigation of IBR-driven instabilities. Specifically, we will use the metrics from Section 2 to quantify the necessary stability margin improvements and develop novel control designs for IBRs that can improve these metrics. We propose to use the control-theoretic concept of model matching, which allows the designing of controllers that can reshape the power grid response. This concept will allow to use flexibility of IBRs for voltage shaping, i.e., to proactively shape voltage response and prevent either small-signal instability (such as sub-synchronous oscillations) or large voltage dips due to weak grid conditions (short circuit ratios of 1 to 3), which is of interest to our Project Affiliates Electric Power Research Institute (EPRI) and DTU. We will further explore combining such methods with the frequency counterpart (frequency shaping), wherein controllers are designed to prevent frequency fluctuations from synchronous resources. Critical to this task will be understanding the tradeoffs between using IBRs for voltage shaping, frequency shaping, or both. The outcome of this task will be a set of control designs with guaranteed improvements in both frequency and voltage response.

Section 4: Integration of Stability Analysis with Operations and Services

The outcomes of Sections 1-3 must be integrated into power grid operation and provision of ancillary services. We will develop a framework wherein the power grid operator dispatch – in addition to determining the reference set-points for IBRs – will co-optimize their control response as necessary to ensure small-signal stability and voltage support (voltage shaping), and frequency support (frequency shaping in Section 3) from IBRs. By integrating stability and frequency needs into dispatch, the system operator will be able to trade-off between the two, prioritizing the response between the time of the primary source and the energy content. Our goal is to use this trading off capability to devise ancillary services, i.e., a set of predefined services that the operator will contract from market participants to maintain a stable, secure, and economical power grid operation. Specifically, we will define services needed to replace fossil-fuel generators with IBRs. We envision that the necessary subset of such services includes: (a) inertia/frequency support services—required to prevent large frequency fluctuations in the absence of fossil-fuel generators; (b) synchronization services—that allow IBR to synchronize with/from fossil-fuel generator frequency; (c) voltage support services—that can mitigate pronounced voltage dips during faults due to the reduced short circuit currents; and (d) restoration services needed to re-energize T100RE power grids and reliable re-connection of demand after large outages (“black start”). The outcome of this task will include resource-dependent stability constraints that can be integrated within power grid dispatch and a list of services needed to support IBR integration. Furthermore, integration of IBRs into stability analysis of T100RE will produce knowledge and tools for T100RE stakeholders to analyze and mitigate IBR-driven instabilities, providing early warning and effectively restoring an IBR-dominated power grid. These outcomes will facilitate the removal of constraints on operating inverter-based VREs and their more efficient utilization.