Europe’s electricity system is undergoing a significant transformation as conventional synchronous generation is increasingly replaced by inverter-based resources (IBRs). While this transition supports decarbonisation, it also introduces new dynamic behaviours that are not adequately addressed by existing grid connection requirements. One emerging concern is forced oscillations – periodic oscillations in active power caused by external disturbances such as wind turbulence, wave loading, tower shadow, wind shear and blade-passing frequency.

Unlike naturally occurring electromechanical oscillations within the grid, forced oscillations originate from external periodic sources. If their frequencies coincide with the natural oscillatory modes of the power system, resonance may occur, significantly amplifying oscillation amplitudes. Such resonance can reduce system stability, damage equipment, trigger protection systems, and in extreme cases, contribute to widespread system disturbances. Existing European grid codes limit harmonic current injection above the fundamental frequency but contain no explicit provisions for oscillations in the 0.1–20 hertz (Hz) range below the fundamental frequency. The European Commission’s (EC) draft Network Code on Requirements for Generators (NC RfG) 2.0 introduces new limits, specifically targeting this frequency band.

Against this background, in July 2026, the European Network of Transmission System Operators for Electricity (ENTSO-E) and WindEurope, jointly released the Technical Group on Forced Oscillations Report, to provide technical guidance on the proposed forced oscillation requirements introduced in the draft NC RfG 2.0. Although the report is non-binding, it is intended to form the technical basis for a future implementation guidance document after the EC formally adopts the revised network code. The report primarily focuses on wind power plants, although future requirements will also apply to other inverter-based power park modules (PPMs), including solar photovoltaic (PV) plants and battery energy storage systems (BESS).

According to the report, wind turbines naturally exhibit structural oscillations due to wind loading, wave excitation and tower dynamics, while active tower damping systems deployed to mitigate them can also induce oscillations in active power. The cumulative impact of these oscillations on the future electricity system with much higher shares of IBRs remains uncertain. Consequently, minimum performance requirements are considered necessary to ensure system security while avoiding unnecessary design burdens on renewable energy projects. Notably, some ENTSO-E members of the Expert Panel on the April 2025 Iberian blackout also reviewed this report to ensure consistency with its findings – an indication of Europe’s increasing focus on grid stability as inverter-based generation becomes more widespread.

Regulatory context

The report originates from the amendment process initiated by the EC in 2022 to modernise the three European Connection Network Codes (CNC) – RfG, demand connection and high voltage direct current (HVDC). Following stakeholder consultations, ACER recommended amendments to NC RfG in December 2023, including two new provisions – Articles 21.3 for onshore wind farms and 26.2 for offshore ones – which, for the first time, introduce limits on forced oscillations in active power.

Recognising the technical uncertainty surrounding these new provisions, ENTSO-E and WindEurope established a joint technical group comprising 12 experts in June 2024. Their objective was not to redefine the legal requirements but to clarify what the draft clauses actually require, pressure-test the proposed ranges against real operating data, and settle on a common measurement methodology. Their latest report represents the outcome of this collaborative work and is intended to guide future implementation across Europe. The EC is expected to adopt a final text sometime in 2026, after which transmission system operators (TSOs) will have up to three years to transpose the requirements nationally.

The new requirements follow a step-by-step approach to limiting forced oscillations. Clauses a(i) and a(ii) set the maximum allowable amplitude of sustained oscillations, expressed as a percentage of either the turbine’s actual active power output or its rated capacity. If these limits are exceeded, clauses b(i), b(ii) and b(iii) apply. Clause b(i) limits the maximum amplitude of a temporary exceedance, b(ii) limits how long the exceedance may persist, and b(iii) requires the oscillation to decay to half of its peak value within a specified time. Clauses c(i) and c(ii) then limit how frequently such exceedances can occur by setting caps on the percentage of time they are permitted each day and the maximum number of events allowed per hour. Each clause specifies a most strict, default and least strict threshold, giving TSOs flexibility in selecting the applicable value. One of the technical group’s key tasks was to assess which of these threshold values are practical for real-world wind farm operation.

The proposed requirements distinguish between onshore and offshore wind farms (OWFs), with offshore projects receiving somewhat more relaxed limits in the 0.1–2 Hz frequency range because larger offshore turbines are inherently more susceptible to structural oscillations and mitigation measures are more costly.

Assessment methodology

A major contribution of the report is the development of a non-binding methodology for assessing compliance with the proposed forced oscillation limits. The methodology consists of four principal stages: preprocessing the measurement data; filtering the signal to isolate the frequency range of interest; detecting oscillatory events; and classifying whether detected oscillations violate the proposed NC RfG limits.

The input data is typically one week of active power measurements sampled at 10 Hz or higher. Missing data may be interpolated, while smoothing is generally discouraged because it may suppress genuine oscillations. The methodology employs a sixth-order Butterworth band-pass filter to isolate oscillations, typically within the 0.1–2 Hz range for standard monitoring applications.

After filtering, the algorithm identifies local maxima and compares them against dynamic amplitude limits defined in Articles 21.3 and 26.2. Consecutive exceedances are grouped into “violation envelopes”, allowing the algorithm to determine their peak amplitude, duration and frequency of occurrence. These detected events are then evaluated against the different compliance clauses governing instantaneous amplitude, duration, damping and event frequency.

The report also presents a prototype MATLAB implementation of this methodology. Although not publicly released, it served as the common analytical platform for all technical group members when evaluating real operational data from existing wind farms.

The technical group also examined a more sophisticated alternative, the continuous wavelet transformation (CWT), which handled multi-frequency oscillations more elegantly than simple peak-detection, but shelved it for now over uncertainty about pending patents on the technique, preferring not to give a compliance advantage to developers willing to license a patented algorithm.

Grid code compliance recommendations

The report provides extensive guidance on how TSOs and generators should demonstrate compliance. Measurements should generally be taken at the point of common coupling (PCC), although alternative measurement locations may be agreed upon where multiple connection points exist, particularly for OWFs connected through artificial islands or offshore hubs. Measurement equipment should comply with IEC/IEEE PMU standards, preferably using M-class phasor measurement units (PMUs) with suitable accuracy.

Sampling requirements depend on the frequency range being assessed. A 100 ms resolution (10 Hz) is considered sufficient for monitoring forced oscillations up to 2 Hz, while assessment up to 20 Hz requires higher sampling rates of around 50 Hz (20 ms). Beyond 2 Hz, the report recommends that the sampling rate be agreed on a project-specific basis between the grid user and the relevant TSO. Data should be retained for one year and exchanged securely between the grid user and the TSO.

Compliance testing should begin only after at least two months of full commercial operation. If exceedances are detected, the report recommends a cooperative process in which TSOs and developers investigate mitigation measures, implement corrective actions and repeat measurements before any regulatory consequences are considered. Importantly, oscillations caused during legitimate system support services, including fault ride-through, frequency control, synthetic inertia and other mandatory grid-support functions, should be excluded from compliance assessment.

The report further recommends monitoring over a continuous one-year period to capture seasonal variations in wind and wave conditions, rather than drawing conclusions from only a few weeks of data.

Assessment of existing wind farms

The report analyses operational data from existing European wind farms to evaluate how realistic the proposed limits are. Three independent datasets were examined:

  • Data supplied by one wind turbine manufacturer (OEM1) covering five onshore wind farms;
  • Data from a second manufacturer (OEM2) covering nine sites;
  • Operational data from an offshore wind developer (DEV1) covering three large OWFs.

Onshore: The OEM1 analysis revealed that the least strict and default thresholds generally produced very high compliance rates. Most failures occurred under clause b(i), which limits instantaneous oscillation amplitude. Detailed investigation showed that many of these apparent violations were not genuine forced oscillations but resulted from legitimate operational events such as curtailment removal and rapid power set-point changes imposed by the grid operator. Once these events were recognised, the technical group suggested that they should be excluded from the analysis of forced oscillations.

The study also found that the strictest limits were largely impractical, producing frequent non-compliance despite otherwise normal turbine operation. While oscillations generally decayed rapidly enough to meet the damping requirements, the instantaneous-amplitude clause b(i) remained the single biggest source of failure under the strictest thresholds, compounded by how often exceedances recurred.

The OEM2 analysis reached similar conclusions. Under default limits, several sites approached or exceeded the proposed thresholds, whereas applying the strictest values caused seven of nine analysed sites to fail. This reinforced the view that the strictest limits would be unrealistic for existing commercial wind farms.

Offshore: The offshore assessment produced somewhat different results. Because offshore projects are allowed more generous thresholds, all analysed wind farms satisfied the least strict requirements. Under default thresholds, around four-fifths of the analysed weeks remained compliant. Again, the strictest limits proved overly restrictive, with compliance falling dramatically.

The report attributes the generally better offshore performance partly to higher allowable limits and partly to the different operating characteristics of OWFs. Importantly, even when oscillations exceeded default thresholds, they typically decayed quickly enough to satisfy the damping requirements, suggesting that the primary challenge lies in the number of exceedance events rather than persistent oscillatory behaviour.

Performance of wind farms under different threshold scenarios

Sensitivity analyses

Sensitivity analyses showed that compliance outcomes were most sensitive to the continuous amplitude limits in clauses a(i) and a(ii), with changes to these thresholds having a much greater impact than the frequency-of-recurrence limits in clauses c(i) and c(ii). Relaxing the continuous amplitude thresholds substantially improved compliance across the analysed wind farms, whereas the instantaneous-amplitude limit in clause b(i) remained the principal source of non-compliance. The technical group attributed many of these exceedances to legitimate operational events, such as power set-point changes, and, therefore, recommended introducing a minimum-duration filter to exclude such events from compliance assessment. The report concludes that the continuous amplitude limits should be carefully calibrated as the primary regulatory lever, while noting that it has not independently validated the grid-stability risk case for any particular threshold.

Mitigation measures

Mitigation of forced oscillations in wind farms involves a combination of electrical and structural solutions, chosen based on the specific technical and economic context.

Electrical mitigation options focus on modifying the electrical system to absorb, dissipate, or control oscillations. These include:

  • Energy storage systems (mechanical, electrochemical, electrical, thermal) that can absorb excess energy during oscillations and release it when needed, helping to smooth power output
  • Dissipation (load banks): Devices that absorb excess power as heat, reducing oscillation amplitude
  • Demand side management to adjust demand in response to oscillations to help stabilise the grid
  • Flexible AC transmission (FACT) technologies such as thyristor-controlled series capacitor(TCSC), static synchronous compensator (STATCOM), static synchronous series compensator (SSSC) and unified power flow controller (UPFC) that can dynamically control voltage, impedance and phase angle to dampen oscillations; and
  • HVDC links can be used to control power flows and dampen oscillations between grid regions.

These solutions can be implemented at the wind turbine generator, wind farm, or grid level, depending on the scale and nature of the oscillation problem.

Structural solutions are typically implemented at the turbine or wind farm level and are especially relevant for mitigating oscillations caused by wind, waves, or structural resonances. Because many of these oscillations originate from unavoidable environmental loading such as wind and waves, most mitigation options work by managing their electrical signature rather than eliminating the mechanical motion itself – the report does not prescribe a preferred approach at this stage. It encourages flexibility and ongoing innovation, with a focus on practical, case-specific measures.

Evidence gaps above 2 Hz and WindEurope’s dissent

The report acknowledges that the available evidence remains limited, particularly for oscillations above 2 Hz. As most of the onshore data was sampled at 10 Hz, the compliance assessment focused mainly on the 0.1-2 Hz frequency range. A smaller offshore dataset with higher-resolution measurements found no compliance violations between 2 Hz and 4 Hz. However, Denmark’s TSO, Energinet, reported sustained 3 Hz oscillations in power electronics-dominated parts of its grid, highlighting the need for further monitoring and analysis of higher-frequency oscillations covered by the draft NC RfG 2.0.

WindEurope also issued a formal dissent, arguing that the evidence does not yet justify the proposed limits. It cautions that overly stringent thresholds could lead to unnecessary turbine redesigns, higher project costs and slower wind energy deployment. The association recommends treating the current default values as the maximum permissible limits, with any stricter requirements needing technical justification from the relevant TSO. It also argues that the compliance requirements should apply only to very large PPMs rather than all projects.

Main conclusions

The report concludes that introducing explicit limits on forced oscillations represents an important evolution of European grid codes as IBRs become increasingly dominant. However, successful implementation requires practical assessment methods, realistic compliance thresholds and close cooperation between TSOs and the wind industry.

The proposed methodology provides a consistent framework for identifying and assessing oscillatory behaviour, while the analysis of operational wind farm data demonstrates that default and least strict thresholds are generally achievable, whereas the strictest limits are not compatible with the normal operation of many existing wind farms.

The technical group therefore recommends adopting balanced threshold values that maintain system security without imposing disproportionate costs on renewable energy developers. It also recommends continued monitoring, further assessment of solar PV and BESS technologies, and future refinement of detection algorithms as operational experience accumulates.

Net net, as Europe moves towards an increasingly inverter-based power system, the report provides a practical framework for managing forced oscillations while highlighting the need for continued research and collaboration.