Executive Summary
Conventional turbine-generator shafts have very low inherent torsional damping. When exposed to certain grid conditions—especially with high levels of inverter-based resources (IBRs) and series-compensated transmission—these shafts can experience sub-synchronous oscillations (SSOs). Two families matter most: sub-synchronous resonance (SSR), classically driven by network resonance near a generator’s torsional modes, and sub-synchronous control interaction (SSCI), in which power-electronic controls (e.g., wind plants, STATCOMs, or batteries) feed negative damping into those modes. The industry has learned hard lessons—from the Mohave generator shaft failures in the 1970s to the 2009 Texas SSCI incidents—and modern grids with rising IBR penetration require updated study methods, models, and mitigations that extend beyond traditional synchronous-centric practices (ERCOT), (ESIG), (EPRI).
This post explains torsional phenomena in plain language, clarifies how IBR performance deficiencies can exacerbate risk, and outlines pragmatic, forward-leaning mitigations—particularly the role of grid-forming (GFM) BESS. The takeaway is realistic and positive: we can continue moving toward very high penetrations of wind, solar, and storage while protecting conventional generators by (1) tightening IBR performance and model quality, (2) adopting GFM controls with proven stability features, (3) modernizing studies (EMT/co-simulation) and plant tests, and (4) deploying targeted protections (TSR relays) and network mitigations where needed (NERC), (MISO), (PNNL), (GE Vernova).
1. Foundations
1.1 What “torsional effects” are and why they matter
Torsional vibration is the periodic twisting of a generator shaft about its axis. A multi-mass shaft train (HP/IP/LP turbine sections plus the generator) has several natural torsional frequencies—like a string with multiple harmonics. If the grid pushes at one of these frequencies, the twisting can amplify dramatically (TAMU).
Plain-language translation: imagine holding a long, slightly springy ruler and twisting it. If someone taps it rhythmically at just the wrong pace, the ruler starts to twist more and more. That “just the wrong pace” is the natural frequency; grid currents can tap at that pace.
Real-world example: in the Mohave coal plant, series-compensated lines created an electrical resonance that matched shaft modes; the generator shafts failed in 1970 and 1971, which cemented SSR as a reliability concern in classical, synchronous-dominated grids (ERCOT).
1.2 From SSR to SSCI: how IBRs enter the picture
- SSR (sub-synchronous resonance): electrical resonance (often from series capacitors) interacts with a generator shaft mode, supplying energy near the torsional frequency.
- SSCI (sub-synchronous control interaction): power-electronic controls—including those in IBRs—effectively inject negative damping into sub-synchronous frequencies, energizing shaft modes without a strong network resonance. SSCI was observed in wind plants on series-compensated lines in Texas around 2009 (ERCOT), (Electranix), (IPST).
Plain-language translation: SSR is like pushing a child on a swing because the playground structure happens to bounce at just the right rhythm; SSCI is like a motorized pusher that “learns” the wrong rhythm and keeps adding energy to the swing.
1.3 Why this topic is resurging with high IBR grids
IBR penetration is rising rapidly. NERC has issued multiple alerts about IBR performance and model-quality issues; many events have involved unexpected tripping or poor dynamic behavior during disturbances. Poorly damped IBR controls can aggravate sub-synchronous modes or trigger new ones—especially when combined with series compensation or atypical grid strength (NERC), (NYSRC), (MISO).
Put simply, we’re adding millions of agile but very fast devices (IBRs) to a system originally designed for slower, heavy machines (synchronous generators). If some of the agile devices don’t follow the “dance steps” precisely—especially during disturbances—they can accidentally pump energy into the twisting of the big machine shafts.
1.4 Who pioneered solutions and today’s leadership
Decades of SSR research came from utilities, OEMs, EPRI, and universities; modern SSCI work adds IBR manufacturers, national labs, and system operators. Today’s leadership includes ESIG (practice guidance), NREL/UNIFI and DOE (GFM research and specifications), WECC/PNNL and others (GFM models), and vendors such as GE Vernova (torsional protection) (ESIG), (DOE), (NREL), (PNNL), (GE Vernova).
Summary: Torsional issues are old, but their causes are evolving. IBR controls can interact with synchronous shafts in new ways. The industry is actively building the models, specs, and protections to keep moving toward high IBR penetration safely.
2. Core Concepts & Mechanics
2.1 The multi-mass shaft and its modes
A turbine-generator train can be simplified as several inertias linked by torsional springs and dampers. Each “mass-spring” pair produces natural frequencies \( \omega_{n} \) and damping ratios \( \zeta \). A lightly damped mode near a grid-excited frequency is vulnerable.
Plain-language translation: think of several flywheels connected by springy couplings. Twisting one wheel excites a wobble that sloshes through the chain at specific “favorite” rhythms.
Real-world illustration: OEMs often identify torsional modes via strain-gauge testing or encoder-based twist measurements during start-ups or controlled ramps, yielding mode shapes and frequencies for study and protection setting (ABB), (Kelm).
2.2 From network impedance to control interactions
In SSR, the key is network resonance created by series capacitors: the Thevenin equivalent as seen from the generator terminals exhibits a sub-synchronous resonant frequency that can supply energy to a shaft mode. In SSCI, the network may be strong or weak, but IBR control loops, PLLs, or current regulators interact with series compensation or grid impedance in ways that create an effective negative damping at sub-synchronous frequencies (EPRI), (Electranix), (IPST).
Plain-language translation: SSR is the grid “ringing” at a bad frequency; SSCI is a controller “pushing” at a bad frequency.
Analogy: picture a dancer (IBR) following a beat. If the dancer’s timing algorithm glances at the room echo (series capacitor resonance) and decides to lead, not follow, at a slightly off beat, the dancer and the echo can reinforce each other and shake the stage.
2.3 A minimal math view (kept intuitive)
Consider one torsional mode represented by: \( J \ddot{\theta} + D \dot{\theta} + K \theta = T(t), \) where \( \theta \) is twist angle, \( J \) inertia, \( D \) damping, \( K \) torsional stiffness, and \( T(t) \) the electrical torque perturbation.
- Natural frequency: \( \omega_n = \sqrt{K/J} \).
- Damping ratio: \( \zeta = \tfrac{D}{2 \sqrt{JK}} \).
- If the grid/control injects torque at \( \omega \approx \omega_n \) and the effective damping becomes negative (i.e., the control contributes \( -D_\text{eff} \)), the oscillation grows.
Plain-language translation: the more inertia and stiffness, the faster the small twists vibrate; if someone pushes right at that speed and your brakes (damping) are too weak—or worse, someone is secretly stepping on the “anti-brake”—the motion can run away.
Real-world heuristic: many dangerous torsional modes sit between ~10–55 Hz for large multi-mass trains; series compensation percentages and IBR controller bandwidths can line up here. Screening for “reactance crossover” and eigenvalue damping provides early flags (IPST), (Electranix).
2.4 What makes IBRs different from synchronous machines at sub-synchronous frequencies
Synchronous machines inherently follow the mechanical angle; at sub-synchronous frequencies, their electromechanical coupling behaves in predictable ways shaped by machine physics. IBRs, by contrast, are software-defined: PLLs, current limits, voltage control, and active damping filters dictate behavior. This flexibility is powerful but also means that poorly tuned controls or incomplete models can accidentally provide negative damping at certain frequencies (NERC), (ESIG).
Plain-language translation: synchronous machines are “analog” dancers with fixed instincts; IBRs are “digital” dancers who can dance any step—but if they learn the wrong step, they can stomp on someone’s foot.
2.5 Where grid-forming (GFM) fits
Grid-forming inverters maintain an internal voltage source and share load without relying on a phase-locked loop. Properly designed, they can behave more like “virtual synchronous machines” and offer robust damping and ride-through—even in weak grids. Research and field experience indicate GFM can improve stability margins and reduce harmful interactions when current limiting, droop settings, and inner-loop dynamics are designed explicitly for network interactions (ESIG), (NREL), (DOE), (PNNL).
Plain-language translation: a GFM battery acts like a calm metronome rather than a follower searching for the beat; that steadiness can keep everyone else in time.
Real-world example: field tests (e.g., Hornsdale Power Reserve) and lab demonstrations show GFM units supporting black start, fault ride-through, and damping in weak systems—capabilities now moving into specifications through UNIFI and emerging models such as WECC’s REGFM_A1 (ESIG), (NREL), (PNNL).
Summary: Torsional issues arise when network or control dynamics inject energy into lightly damped shaft modes. GFM offers a pathway to shape those dynamics constructively, provided models and settings are right.
3. Applications & Implications
3.1 Situations with elevated torsional risk in high-IBR systems
- 3.1.1 Series-compensated corridors with nearby conventional units and wind/solar/BESS: classic SSR risk plus SSCI potential. Texas events demonstrated SSCI from DFIG wind plants interacting with series capacitors (ERCOT), (IPST).
- 3.1.2 Weak-grid pockets: low short-circuit strength elevates the influence of controls and can shift resonances; IBRs with aggressive current regulators or poorly tuned PLLs can destabilize sub-synchronous modes (ESIG), (NERC).
- 3.1.3 Mixed fleets with STATCOMs/HVDC/FACTS: power-electronics devices near conventional plants can cause torsional interactions if their control filters align with shaft modes, even without series caps (EPRI).
- 3.1.4 During outages or reconfigurations: topology changes can inadvertently align impedances and control bandwidths with shaft modes (NASPI).
Plain-language implication: any time you add fast, strong electronics near a big spinning shaft, check the rhythms.
3.2 What recent reliability alerts and requirements mean
NERC’s alerts on IBR performance and model quality require owners to review protection settings, ride-through behavior, control configurations, and to submit and validate models that capture relevant dynamics. Regions like MISO and the NYSRC explicitly reference IEEE 2800-based performance requirements and stronger modeling and measurement obligations (NERC), (MISO), (NYSRC).
Why it matters: credible EMT-ready models of IBRs and their control modes are now table stakes for accurate SSO screening and detailed studies; without them, torsional risk can be mis-estimated.
3.3 Study workflow that actually works (and moves projects forward)
A pragmatic, pro-deployment workflow combines speed and rigor:
- Screen with frequency-domain tools: use eigenvalue damping and reactance-crossover methods to flag sub-synchronic modes and potential negative damping tendencies with and without series caps (IPST), (Electranix).
- Run EMT for high-risk scenarios: include full IBR controls (grid-following and grid-forming), current limiters, PLL (if present), droop blocks, active damping filters, and any FACTS/HVDC controls. Co-simulate transmission and distribution where DERs materially alter result; modern toolchains stitch PSS®E + PSCAD + OpenDSS for this purpose (NREL).
- Validate with measurements: plant testing, PMU-based oscillation monitoring, and torsional sensing on generators (strain-gauge or encoder twist) confirm actual mode frequencies and damping, and help tune protection (NASPI), (Kelm).
- Iterate settings/models: update droop, current limits, PLL bandwidth (or eliminate PLLs with GFM), and add supplemental damping as needed. Capture changes in the as-built model set per regional requirements (MISO), (NYSRC).
Plain-language translation: first, check the map; then drive the road in a simulator; then measure real traffic and adjust your route.
3.4 The constructive role of GFM BESS
Properly engineered GFM BESS can:
- Provide sub-synchronous damping: by acting as a stiff, well-damped voltage source with tuned virtual impedance and current limiting, GFM units reduce negative damping pathways.
- Improve fault ride-through and recovery: maintaining controllability through faults reduces control hunting that can feed oscillations.
- Enable black start/“grid-forming pockets”: a stable local reference prevents PLL chasing and restores order after disturbances (ESIG), (DOE), (NREL), (PNNL).
Caveat: GFM is not a magic wand. Current limiting transitions, outer-loop droop values, and inner-loop filter choices must be designed to avoid inadvertently exciting sub-synchronous bands. UNIFI specifications and emerging WECC models (REGFM_A1) are steps toward consistent, operator-ready behavior (DOE), (NREL), (PNNL).
3.5 Protections and mitigations for conventional units
- Torsional stress relays (TSR): modern relays monitor shaft torque and accumulated fatigue; they can alarm or trip if thresholds are exceeded, providing a last-line defense (GE Vernova).
- FACTS tuning and damping controls: retune series-cap bypass strategies, add damping controllers, or reduce compensation when SSO risk is high (EPRI).
- Plant retuning: adjust exciter/PSS, turbine controls, or clutching strategies to alter modal damping.
- Network reconfiguration: during weak or stressed system conditions, temporarily alter flows or compensation to avoid dangerous alignments (NASPI).
Plain-language translation: put better seatbelts on the generator (TSR), teach the surrounding devices to be calmer dancers (tuning), and if the dance floor gets too wild, change the playlist (network config).
3.6 Case sketches
- ERCOT SSCI experience: Damage to series capacitors and wind equipment circa 2009 led to systematic SSCI study methods and criteria; many regions now expect damping to recover within seconds and see EMT studies as mandatory in high-risk corridors (ERCOT), (Electranix).
- OEM torsional protection programs: Utilities deploying TSRs on units near heavily compensated lines and dense IBR clusters report actionable alarms and trip evidence during events—supporting both post-mortems and retuning (GE Vernova).
Summary: With the right study stack, GFM-centric designs, and targeted protections, we can keep onboarding IBRs aggressively while safeguarding conventional shafts.
4. Integration & Broader Context
4.1 How this ties into adjacent domains
- Control theory & power electronics: SSCI is fundamentally a control-interaction problem; using loop-shaping, virtual impedance, and robust current-limiting transitions is key for GFM (NREL), (PNNL).
- Mechanics & materials: cumulative torsional fatigue accumulates even if no single event trips a unit—TSR data help quantify life consumption and inform maintenance (GE Vernova).
- Planning economics: series compensation boosts corridor transfer cheaply, but may require added damping systems, monitoring, or BESS siting to avoid SSO—costs that should be internalized in planning (EPRI).
- Protection & operations: PMU-based oscillation monitoring and plant-level tests are becoming part of “steady-state” operations in high-IBR grids; operators need playbooks for topology-dependent SSO risk (NASPI).
4.2 From theory to practice: a “design-to-operate” chain
Code → Controls → Validation → Operations should become standard for IBR/thermal hybrids:
- 4.2.1 Controls: pick GFM schemes with explicit sub-synchronous damping (e.g., droop and virtual impedance targets that avoid shaft modes).
- 4.2.2 Models: deliver EMT-ready models (REGFM_A1 or equivalent) with current limiters, saturation, and grid-strength sensitivity built in (PNNL), (NREL).
- 4.2.3 Validation: laboratory HIL and staged field tests to confirm oscillation damping and current limiting under faults and weak grids (DOE), (NREL).
- 4.3.4 Operations: establish real-time oscillation detection and TSR alarms; maintain a library of “high-risk topologies” and pre-planned mitigations (NASPI), (GE Vernova).
4.3 Open questions and research frontiers
- Standardizing GFM “fault-behavior envelopes”: industry consensus on current-limit transitions and ride-through that guarantee no negative damping across wide grid strengths (DOE), (NREL).
- Scalable co-simulation & digital twins: faster EMT plus T&D co-simulation to make SSO risk assessment routine in interconnection queues (NREL).
- Online damping estimation: using ambient PMU data and inverter telemetry to infer real-time sub-synchronous damping margins, enabling proactive re-dispatch or retuning (NASPI).
- Market signals for damping: compensation for BESS providing proven oscillation damping, measured via PMUs/TSR statistics—akin to inertia or fast frequency response products (EPRI).
Summary: The path forward blends control engineering, mechanical fatigue management, advanced modeling, and operational monitoring—turning SSO risk into a manageable engineering constraint rather than a blocker to high IBR penetration.
5. Practical Guidance
5.1 Rules of thumb (for planners, OEMs, and developers)
- Always screen SSO when series caps ≥ ~20–30% and a thermal unit is in electrical proximity to IBRs or FACTS; if flagged, escalate to EMT (EPRI), (ERCOT).
- Prefer GFM BESS at or near electrically central buses in weak pockets; tune droop and virtual impedance to avoid 10–55 Hz interactions; verify with EMT and staged tests (ESIG), (NREL).
- Demand EMT-complete vendor models for grid-forming and grid-following modes—including current limiters, PLLs (if any), and ride-through logic—and validate against site tests (NERC), (MISO), (NYSRC).
- Instrument at least one nearby thermal unit with TSR and integrate PMU-based oscillation monitoring; use data to refine mode maps and protection thresholds (GE Vernova), (NASPI).
- Treat series compensation as dynamic: plan bypass switching, damping control, or alternative transfer upgrades (e.g., strategically placed BESS/STATCOM) in corridors with SSO history (EPRI).
5.2 A minimal, modern study package for interconnection
- 5.2.1 Frequency-domain screen across scenarios (N-1, seasonal flows) with varying cap levels and dispatch.
- 5.2.2 EMT cases for worst alignments with full IBR controls and protection models; include GFM variants.
- 5.2.3 Vendor-witnessed staged tests to calibrate models: frequency sweeps (where allowable), fault ride-through under weak conditions, and oscillation damping checks.
- 5.2.4 Protection review: TSR settings, plant tripping logic, wide-area monitoring alarms.
- 5.2.5 Operational playbook: topology-dependent risk tables and remedial action (Electranix), (NERC), (NASPI).
5.3 GFM-specific design pointers
- Current limiting must be smooth and grid-compatible; abrupt mode changes can excite sub-synchronic bands.
- Avoid overly tight droop with slow outer-loop filters; ensure phase/gain margins across expected grid strengths.
- Use virtual impedance to shape sub-synchronous behavior; verify that zero-sequence and negative-sequence paths don’t reintroduce issues.
- Provide telemetry for online damping estimation and event forensics (NREL), (PNNL), (ESIG).
5.4 Communicating with stakeholders
For operators and regulators, frame GFM as a reliability enabler: better dynamic performance and controllability in weak grids, with clear lab and field evidence. For thermal fleet owners, emphasize protections (TSR), credible models, and staged testing that de-risk operations while preserving transfer capability (GE Vernova), (DOE), (ESIG).
Summary: A disciplined but deployment-friendly package—screening → EMT → validation → protections—lets us place GFM BESS strategically, set robust requirements for all IBRs, and keep conventional units safe as we accelerate the transition.
6. Appendices—Short Mental Models
- “The Metronome and the Dancers”: A GFM battery is the metronome; other devices follow. When the metronome is solid and well-damped, the dance stays orderly.
- “Seatbelts and Speed Limits”: TSRs are seatbelts; damping requirements and current-limit envelopes are speed limits. Use both.
- “Map, Drive, Measure”: Screen the map (frequency domain), drive it in the simulator (EMT), and measure on the road (PMUs/TSR). Repeat.
Select Citations
- (EPRI) ASSESSING RISK OF SUB-SYNCHRONOUS OSCILLATIONS
- (NERC) Inverter-Based Resource Performance Issues: Public Report (2023) and Alerts
- (ESIG) Grid-Forming Technology in Energy Systems Integration (2022) and Diagnosis and Mitigation of Observed Oscillations in IBRs (2024)
- (ERCOT) Sub-Synchronous Control Instability Studies: IBRWG (Oct 11, 2024)
- (GE Vernova) Sub-Synchronous Torsional Interaction Analysis and Torsional Stress Relay
- (Electranix) SSCI Detailed Evaluation Using PSCAD (Tech Memo)
- (NASPI) Sub-Synchronous Oscillation Overview (2025 EATT)
- (MISO) IBR Performance Requirements / IEEE 2800 Updates (2024) and Background & Rationale (2023)
- (NYSRC) Guidance to IBR Interconnection Customers on Reliability Rule B.5 (2024) and NERC Alert Attachment via NYSRC
- (NREL) Research Roadmap on Grid-Forming Inverters, Introduction to GFM Inverters (2024), Co-Simulation for IBR Stability (2025), and UNIFI Specifications (2024)
- (PNNL) New Grid-Forming Inverter Models Help Utilities Plan
- (DOE) Powering On with Grid-Forming Inverters, Solar Integration Basics, and Grid Modernization Strategy (2024)
- (ABB) Experimental Verification of Torsional Natural Frequencies
- (Kelm) Field Torsional Vibration Measurement
- (IPST) Analysis and Mitigation of SSCI with Series-Compensated Lines (2023)
- (TAMU) Torsional Dynamics Overview (Lecture Notes)
