Southeast Europe’s renewable-energy market is approaching a more technically demanding phase. Grid access is no longer determined only by available megawatts, completion of substations or compliance with voltage, frequency and fault-ride-through requirements, writes Energy.Clarion.Engineer
The next generation of European connection rules will also examine the quality of the active power injected into the network, including whether wind, solar and battery plants produce repetitive fluctuations capable of disturbing the wider Continental European system.
The emerging requirement concerns forced oscillations: periodic variations in active power caused by mechanical movement, environmental conditions, equipment behaviour or control-system action. ENTSO-E and WindEurope have now developed a common assessment framework intended to support the proposed Network Code on Requirements for Generators 2.0, or NC RfG 2.0.
The revised code has not yet been formally adopted by the European Commission. Its provisions therefore remain proposed rather than binding. Even so, projects now entering front-end engineering, turbine procurement, grid-connection design or financing may still be operating under the new regime once it reaches national legislation.
This is particularly relevant for Southeast Europe. Serbia, Montenegro, Bosnia and Herzegovina, North Macedonia, Albania, Kosovo, Croatia, Romania, Bulgaria, Greece, Hungary and Slovenia form part of, or are directly interconnected with, the Continental Europe synchronous area. A disturbance originating in one national system can interact with electromechanical modes extending across several countries. The commercial border between two balancing zones does not prevent the physical propagation of an oscillation.
Forced oscillations differ from ordinary changes in renewable production. A wind farm naturally increases and reduces output as wind conditions change. A solar plant follows irradiance, while a battery changes its power position in response to dispatch instructions. Forced oscillations are repetitive movements occurring at identifiable frequencies and continuing for long enough to interact with the natural dynamics of the power system.
Wind turbines can produce them through tower shadow, blade-passing frequency, wind shear, rotor behaviour, structural eigenmodes or active tower-damping controls. Mechanical movement in the tower and foundation is translated into electrical output through changes in generator torque. Solar and battery plants face a different mechanism, with potential oscillations originating in inverter controls, plant-level controllers, weak-grid interactions or poor coordination between multiple converter systems.
The systemic risk emerges when the frequency of the injected fluctuation approaches one of the network’s natural modes. The interconnected system can amplify the disturbance through resonance. A relatively small oscillation at one generating facility may then produce a much larger response elsewhere, increasing stress on equipment and contributing to protection operations, generator trips, system separation or, in an extreme event, a blackout.
Continental Europe’s principal inter-area modes are generally found between 0.1 and 1.0 hertz, while local and intra-area modes typically extend from 1.0 to 2.0 hertz. These frequencies matter for Southeast Europe because the region combines large synchronous hydro, coal, gas and nuclear units with a rapidly growing volume of inverter-connected renewable generation.
Long-distance flows pass between Romania and Bulgaria, Greece, the Western Balkans, Hungary and Central Europe. Serbia occupies an important transit position within that structure. Serbia, Montenegro and North Macedonia also cooperate through the SMM control block, including the exchange of balancing energy and reserves. The result is a closely connected operating environment in which a plant’s dynamic performance cannot be assessed solely as a local matter.
The regional system’s recent history underlines that exposure. The Continental Europe separation of 8 January 2021 began with cascading transmission events in the Croatian network. Forced oscillations did not cause the event, but the separation demonstrated the systemic importance of Southeast European corridors and the speed with which a local disturbance can affect the wider European grid.
The proposed forced-oscillation limits would apply to Type C and Type D power park modules. Within Continental Europe, the proposed upper threshold for national classification of Type C facilities is 50 MW, while plants connected at 110 kV or above are generally treated as Type D. National transmission system operators would retain discretion to apply lower capacity thresholds.
This brings much of Southeast Europe’s new renewable pipeline into the potential compliance perimeter. Serbia’s 154 MW Čibuk 2, developed by Masdar and Taaleri, consists of 22 Nordex turbines rated at 7 MW each and shares connection infrastructure with the existing 158 MW Čibuk 1. Enlight Renewable Energy’s 94 MW Pupin project extends the existing Kovačica wind complex. Projects such as Crni Vrh, Jasikovo and Montenegro’s Gvozd development, together with major wind investments in Romania, Bulgaria, Croatia and Greece, belong to the same class of assets for which oscillation performance could become a formal condition of continued grid operation.
Shared grid connections create an especially important technical and contractual issue. Oscillations produced by individual turbines or separate wind-farm sections are not necessarily aligned in phase. At an aggregated connection point, part of the fluctuation may cancel out. A project that appears non-compliant when assessed separately can therefore produce an acceptable result when measured as part of a larger interconnected facility.
This is directly relevant to projects sharing substations, switchyards or export lines. The result of the compliance test may depend on whether measurement is performed at the individual plant transformer, the common 110 kV switchyard, the transmission-system access point or an aggregated point of common coupling.
The measurement boundary must be agreed before detailed design is frozen. It should also be reflected in the grid-connection agreement, EPC contract, turbine-supply agreement and operational procedures. Where the generator finances connection facilities that ultimately become TSO property, ownership alone does not determine who must install the monitoring equipment, maintain the data platform or finance mitigation following a failed assessment.
Under the proposed default values for onshore wind farms, continuous forced oscillations would be limited to the higher of ±0.5 per cent of maximum plant capacity or 500 kW. Temporary oscillations could reach ±2.5 per cent of maximum capacity, but output would have to return within the continuous limit in 180 seconds and fall below half of the temporary limit within half that period.
Temporary exceedances would generally be permitted for no more than 1 per cent of each day. The default frequency test would allow no more than three exceedances per hour, assessed at the 95th percentile across the measured week.
The choice of threshold materially changes the compliance result. Testing undertaken for the new methodology found that only 19 per cent of the assessed weeks complied with the strictest temporary onshore limit when a one-second minimum detection period was used. Compliance increased to 87 per cent under the proposed default value. Applying an offshore-style temporary threshold of 4 per cent to the same onshore data lifted the result to 98 per cent.
These differences show why national implementation cannot be treated as a routine technical formality. A TSO selecting the strictest end of the available range could make a standard turbine configuration appear persistently non-compliant. The resulting corrective work may require controller modification, additional equipment or changes to the operating regime.
The underlying methodology remains incomplete. Most of the available operational data covers only the frequency range between 0.1 and 2.0 hertz, although the proposed legal obligation extends as high as 20 hertz. The analysed wind farms and manufacturers have not been identified, and the assessment tool remains a developmental MATLAB prototype rather than a mature, publicly available compliance platform.
The algorithm can also misclassify normal operational events. Turbine shutdowns, active-power ramps, curtailment instructions, wind turbulence and system-support actions may be interpreted as oscillations. A minimum detection period of approximately one second is proposed to reduce false positives, but some projects may require an adjustment between one and ten seconds.
This makes high-quality event recording essential. Starts, stops, curtailment, frequency-response actions, fault recovery and TSO dispatch instructions must be time-synchronised with active-power measurements. The proposed rules exclude oscillations created by system-support instructions, including requested power-oscillation damping, but the plant owner will have to demonstrate that a flagged event qualifies for that exclusion.
Many existing renewable plants do not possess sufficient data resolution. Conventional SCADA historians frequently store active-power measurements at intervals of one second, ten seconds or longer. Such records cannot reliably reconstruct oscillations occurring between 2 and 20 hertz.
The proposed measurement architecture is closer to transmission-system monitoring than ordinary plant SCADA. It calls for a phasor measurement unit, preferably an M-class PMU compliant with IEC/IEEE 60255-118-1, supported by current and voltage transformers with approximately 0.2(s) accuracy.
A measurement resolution of 100 milliseconds is considered sufficient for oscillations up to 2 hertz. Monitoring the full proposed range to 20 hertz requires at least 40 samples per second, while a practical 50 hertz sampling rate provides one sample every 20 milliseconds.
This specification needs to enter FEED and substation design at an early stage. Suitable CT and VT signals, PMU panels, time synchronisation, secure communications, data storage and cybersecurity interfaces must be reserved before procurement. Retrofitting them after energisation can require new protection-panel work, revised testing and an additional outage.
The plant owner would collect the measurements and make the raw data available to the TSO through secure IT and operational-technology infrastructure. Records could be retained for at least one year, with analysis performed weekly, biweekly or monthly. One or two weeks of data would not be considered representative because oscillation behaviour changes with wind conditions, plant availability and operating state.
The compliance process could continue after the project has entered commercial operation. The initial assessment would take place only after at least two months of full wind-farm operation. A project could receive its Final Operational Notification after completing conventional grid-compliance testing, while forced-oscillation monitoring continues separately.
A failed assessment would trigger investigation, mitigation and a new measurement period. That cycle could be repeated more than once. Where a plant remained non-compliant and the owner no longer proposed credible corrective measures, the TSO could withdraw the Final Operational Notification until compliance was restored.
This introduces an unusual project-finance exposure. A wind farm could complete construction, reach energisation, begin commercial operation and satisfy initial lender conditions while retaining a technical risk capable of restricting its future right to operate.
Financing documents will therefore need to treat forced oscillations as a post-completion compliance obligation. The relevant contracts should allocate responsibility for dynamic modelling, PMU installation, baseline monitoring, data processing, OEM investigation, controller retuning, repeat testing, lost production and physical mitigation.
A generic undertaking that the turbines will comply with the applicable grid code may be insufficient. The national TSO could select project-specific limits after the turbine contract has been signed, while the final compliance result may also depend on measurement location, aggregation and the selected detection algorithm.
The issue reaches into structural engineering. Active tower damping commonly uses generator-torque adjustments to reduce tower movement. Those adjustments can create electrical power oscillations. Limiting them may improve the grid-side result while increasing structural loads on the turbine tower and foundation.
Stricter limits could therefore require additional steel, heavier foundations, revised fatigue calculations or recertification of the turbine configuration. Once a tower and foundation design has been certified—and particularly after commissioning—changes can be expensive and technically difficult.
Potential electrical mitigation includes battery storage, load banks, demand-side response, STATCOMs and other flexible AC transmission technologies. Structural measures include tuned-mass dampers, slosh dampers, generator-torque control and pitch-angle variation. The appropriate response will depend on whether the source lies in the turbine structure, individual converter, plant controller or surrounding grid.
Hybrid projects may have an advantage. A properly controlled BESS could absorb rapid active-power variations from a wind or solar facility. Yet the battery is not automatically a solution. Poorly coordinated inverter controls can introduce another oscillatory mode, particularly at electrically weak connection points. The entire plant must therefore be modelled as one control system rather than as separate wind, solar and storage packages.
Southeast Europe enters this transition with an uneven compliance framework. Serbia, North Macedonia and Bosnia and Herzegovina have made substantial progress in implementing the existing NC RfG. Serbia has adopted significance thresholds and general connection requirements, while EMS conducts functional tests and issues approvals for energisation, temporary connection and permanent connection.
Montenegro remains further behind in formalising its classification thresholds, general requirements and derogation procedures. Albania and Kosovo have incorporated numerous technical obligations into existing grid codes but have not completed the same formal implementation structure.
The wider regional weakness is the absence of authorised independent certifiers. TSOs remain responsible for much of the practical validation, relying on commissioning tests, simulations, manufacturer declarations and project documentation. Adding a data-intensive forced-oscillation procedure could deepen connection bottlenecks where several large wind, solar and storage projects approach commissioning simultaneously.
A fragmented national response would compound the problem. Separate thresholds, algorithms and evidence formats adopted by EMS, CGES, NOSBiH, MEPSO, KOSTT and OST would force turbine and inverter manufacturers to produce country-specific engineering packages. It would also reduce the ability of investors and lenders to use standardised technical due-diligence and completion procedures across regional portfolios.
A coordinated SEE protocol could establish common rules for measurement location, sampling frequency, data format, operational exclusions, monitoring duration, assessment algorithms and the treatment of shared connections. National TSOs would retain authority over system-specific limits, but the underlying evidence and testing process would become predictable.
Grid congestion, connection capacity and slow development of transmission infrastructure remain larger constraints on Southeast Europe’s renewable pipeline. Forced oscillations nevertheless introduce another bankability test at the point where turbine design, substation engineering, plant controls and the continuing right to operate meet.
Projects entering FEED today should be designed to produce the evidence that tomorrow’s grid code may require. The relatively modest cost of PMU-ready measurement architecture, high-resolution data storage and clearly allocated compliance responsibilities is easier to finance during construction than controller redesign, structural modification or repeated testing after commercial operation.
Elevated by Energy.Clarion.Engineer
