Renewable Modelling · Inverter-Based Generation

Characteristics of Inverter-Based Generation

Traditional power systems were built around large synchronous generators — rotating machines whose useful grid-support behaviours come for free from physics. Modern wind and solar connect through power-electronic converters, or inverter-based generation (IBG), and behave differently: the same behaviours are possible, but only when the controls, the current rating, the energy source and the grid code provide them. This guide works through the practical differences one by one — inertia, fault current, voltage support, synchronising torque, damping, frequency control, fault ride-through, reactive power, harmonics and black start — and what each one means when you build an IBG model for an EMTP® dynamic study.

Reading time ≈ 36 min · Synchronous machine vs IBG, difference by difference

Power systems grew up around large synchronous generators (SGs) — coal, gas, hydro and nuclear machines bolted directly to the ac grid. Wind and solar mostly connect a different way: through power-electronic converters, or inverter-based generation (IBG). The headline difference is not simply “conventional versus renewable”. It is deeper. A synchronous machine’s helpful grid behaviours are a by-product of its physics; an IBG only does what its controls are told to do, within the limits of its current rating and energy source. This guide takes the differences one at a time, and each time asks the question that matters for a study: is this service actually present, and within what limits?

This is the opening guide of a nine-part series on modelling inverter-based generation. It lays the foundation by placing the IBG next to the synchronous machine, difference by difference; the parts that follow open up the inverter itself, then the RMS / phasor and EMT modelling methods, and then apply those models to frequency, voltage, small-signal, islanding and control-interaction studies. The single message to carry forward is this: a synchronous machine provides many grid services naturally, through physics; an IBG provides them only if its controls, ratings, energy source and grid-code settings are designed to do so.

One comparison to hold in mind

A synchronous generator is a rotating voltage source coupled directly to the grid. A grid-following (GFL) IBG is normally a controlled current source that follows the grid voltage. A grid-forming (GFM) IBG can behave more like a controlled voltage source — but it still has converter current limits and software-defined behaviour. Almost every difference on this page is a consequence of that one contrast.

Abbreviations used on this page

IBGInverter-based generation (or generator)

SGSynchronous generator

RMSRoot-mean-square (phasor) simulation

EMTElectromagnetic transient (time domain)

EMTP®Electromagnetic Transients Program (an EMT tool)

GFL / GFMGrid-following / grid-forming

VSCVoltage-source converter

PVPhotovoltaic

BESSBattery energy storage system

WTGWind-turbine generator

DFIGDoubly-fed induction generator (Type-3 wind)

PLLPhase-locked loop

ROCOFRate of change of frequency

FFRFast frequency response

LFSM-OLimited frequency-sensitive mode – over-frequency

FRTFault ride-through

LVRT / HVRTLow- / high-voltage ride-through

PODPower oscillation damping

PCC / POIPoint of common coupling / interconnection

SCRShort-circuit ratio (grid strength)

AVRAutomatic voltage regulator

PSSPower system stabiliser

STATCOM / SVCStatic synchronous compensator / static var compensator

ac / dcAlternating / direct current

puPer unit

Key idea

A synchronous generator’s grid support is physics: rotating mass gives inertia, machine flux gives high fault current, the internal voltage behind reactance supports voltage, and the rotor angle gives synchronising and damping torque. None of it needs a controller.
An IBG provides those same services only through control, and only within hard limits — the converter current ceiling (often around \(1.1\) pu, but vendor- and project-dependent), the dc-side energy available, whether the inverter is oversized, and what the grid code demands.
Some IBG responses are not equivalent even when present: a ROCOF-based “synthetic inertia” cannot act at the instant of a trip (ROCOF is still zero), and fault current is capped where a machine would deliver several pu.
For a study the message is simple: do not drop an IBG in as a synchronous-machine equivalent. Model the specific capabilities the study needs, and respect each model’s area of validity — short-circuit ratio, voltage range, current limits, PLL stability.

How to read this page

The page is organised difference by difference. For each item it asks the same four questions: (1) how does a synchronous machine behave? (2) how does an IBG behave? (3) why does the difference matter for the grid? and (4) what does it mean for modelling? A short “Modelling implication” line closes each major difference so the practical consequence is never left implicit.

Key terms used on this page

01Inertia

Kinetic energy stored in rotating mass, released automatically to oppose a change in frequency.

02Synthetic inertia

An emulated inertia-like response from an inverter, driven by measured ROCOF and bounded by energy and current limits — not identical to physical inertia.

03Headroom / curtailment

Operating below available power (curtailment) so the plant keeps upward reserve (headroom); the reserve has an energy / opportunity cost.

04Grid-following (GFL)

A controlled current source that follows the grid voltage angle through a PLL; the common mode today.

05Grid-forming (GFM)

A converter that sets its own voltage and frequency reference, behaving more like a voltage source — still current-limited.

06Synchronising torque

The restoring effect that pulls a machine’s rotor angle back toward the rest of the system.

07Power oscillation damping (POD)

A controller that modulates active or reactive power to reduce system oscillations.

08Fault ride-through (FRT)

Staying connected through a low (LVRT) or high (HVRT) voltage event instead of tripping.

09Momentary cessation

The inverter temporarily stops (or greatly reduces) current while staying connected, then resumes.

10Negative-sequence current

A reverse-phase-order component that appears during unbalanced faults; many IBGs suppress it unless required.

11Harmonic

A voltage or current component at an integer multiple of the fundamental (e.g. 250 Hz is the 5th of 50 Hz).

12Black start

Energising part of the network without an external grid supply, to help restart the system after a blackout.

13Short-circuit ratio (SCR)

Grid short-circuit power at the connection point divided by plant rating; low SCR means a weak grid.

14Voltage-source converter (VSC)

The converter type used by modern IBGs; its harmonic and fault behaviour is set by its controls, not a passive reactance.

Physics versus control — the whole page in one box

Synchronous machine — inertia: physical; fault current: physical; voltage source: physical; synchronising torque: physical; damping: partly physical, partly controlled; reactive support: excitation-limited.

IBG — inertia: control-based if enabled; fault current: current-limited; voltage source: only if grid-forming; synchronising torque: not classical for grid-following; damping: control-based if enabled; reactive support: current-rating and control-mode limited.

Section 1

The big picture

For most of the twentieth century, generation meant a large rotating machine. A synchronous generator is magnetically coupled to the grid and turns at a speed fixed by frequency — on a 50 Hz system a two-pole machine spins at 3000 rpm, a four-pole machine at 1500 rpm — and its behaviour is tied to mechanical inertia, an internal voltage, electromagnetic torque and the rotor angle. Wind and solar mostly connect through inverters instead, and the inverter, not physics, decides how current is injected. So the deeper contrast is not the fuel; it is where the behaviour comes from.

This shift is happening at scale, and in two very different forms. Transmission-connected wind and solar plants now reach many hundreds of megawatts at a single connection point, while rooftop PV arrives as a multitude of small installations — a few kilowatts each — whose aggregate is enormous: rooftop PV in Australia’s National Electricity Market reached about 5 GW by 2017, against a system demand of 25–30 GW. Both forms have to be represented in the dynamic simulations operators rely on, and both behave unlike the synchronous machines they displace.

Physics-dominated versus control-dominated

A synchronous generator behaves the way it does because it is a rotating electromagnetic machine: strip away the controls and it still provides inertia, fault current and a voltage source. An IBG has almost no inherent grid behaviour — every service it offers is created by its control software, and bounded by its current rating and energy source.

Table 1 — The contrast in one view: what each technology brings, and where it comes from.
Aspect	Synchronous generator	Inverter-based generation
Nature of behaviour	Physics-dominated	Control-dominated
Rotating mass / inertia	Stored kinetic energy, coupled to the grid	Usually none directly coupled
Fault current	High, several pu	Limited by semiconductors (\(\approx 1.1\) pu)
Internal voltage	Voltage behind reactance	No inherent internal source
Synchronising & damping	Natural	Emulated, if required
Predictability	Standard, well understood	Depends heavily on control design

Section 2

The central idea: natural versus controlled behaviour

This one idea underlies everything that follows. In a synchronous machine, inertia comes from the rotating mass; fault current from the machine’s electromagnetic flux; voltage support from the internal generated voltage and the excitation system; synchronising torque from the rotor-angle relationship; and damping from the damper windings and the power system stabiliser. None of it is switched on — it is simply how the machine works.

An IBG is different. Its behaviour is not inherent; it depends on a list of design and setting choices: the inverter control algorithm; the converter current limit; the dc-side energy available; whether the inverter is oversized; whether active power can be curtailed; whether storage is present; the grid-code requirements; the PLL and measurement performance; and the voltage during the disturbance. Change any of these and the response changes.

The modelling consequence

Because an IBG’s behaviour lives in its controls, you cannot assume it behaves like a synchronous machine. Two IBGs with identical ratings can respond completely differently. That is why the rest of this guide keeps returning to one question: is this service actually present, and within what limits?

Section 3

Difference 1 — Inertia and synthetic inertia

When generation and demand suddenly diverge — say a large unit trips — the system frequency begins to fall. A synchronous machine responds on its own: its spinning rotor releases kinetic energy, its electrical output rises immediately, and the frequency fall is slowed. This is instant and free, because it is machine physics. An inverter has no directly coupled rotating mass. Even when there is inertia behind it, such as a wind-turbine rotor, that energy is not visible to the grid unless the controls deliberately release it — which is why IBGs are said to have no inherent inertia.

An IBG can emulate inertia. A synthetic-inertia controller measures ROCOF and adjusts active power — if frequency is falling quickly, it pushes more active power out to support the grid. But this is not fully equivalent to machine inertia, because the inverter must measure frequency, compute ROCOF, filter the measurement, act through control delays, respect its current limit, and draw on limited stored energy — and the support may later be withdrawn. Synthetic inertia should therefore not be described as identical to synchronous inertia. In practice it is still being carefully evaluated rather than used routinely, and the short-term overload that would supply it is itself modest — an air-cooled IGBT converter can exceed its rating only for roughly a second.

Why a ROCOF response cannot act at the instant of the trip

At the exact moment a generator trips, the frequency has not yet moved, so ROCOF is still zero.
ROCOF must first be measured or estimated, which needs a time window and filtering.
The controller then changes the active-power reference.
The converter and its primary source must actually deliver the extra power.

A synchronous machine, by contrast, reacts to the power imbalance itself, so its inertia acts immediately. Synthetic inertia is useful — but it cannot be mathematically instantaneous, and it is not the same thing as physical inertia.

Modelling implication. If synthetic inertia or fast frequency response (FFR) is being studied, model the ROCOF measurement, the control delay, the current limit and the available headroom, storage or rotor energy — and the recovery afterwards — not just an idealised inertia constant.

Section 4

Difference 2 — Fault current contribution

During a short circuit, a synchronous generator naturally supplies a large fault current — several times rated — from the flux linkages of the rotating machine (the law of constant flux) and its internal voltage source. Traditional protection was designed around exactly that behaviour. An inverter is limited by its semiconductor current ratings and protection; it usually cannot exceed roughly \(1.1\) pu. To inject reactive current during a fault it typically has to reduce active current so the total stays inside the limit — in the extreme, dropping active power to zero so that all the available current becomes reactive. So during a fault the IBG may cut active power, prioritise reactive current, keep the total current bounded, and behave according to its control strategy rather than machine physics. There is no single universal number for the ceiling: the actual value depends on the vendor design, the control mode, the grid code and the converter rating.

The very-low-voltage problem

If the fault is close to the PCC, the voltage there can collapse to a very low value. When the voltage is extremely low, the inverter may be unable to determine the voltage phase angle, and without a reliable angle it cannot inject current in a controlled way. Many grid codes therefore allow an IBG to stop injecting current below a residual-voltage threshold — often around 20% of rated voltage.

There is an important nuance here. The current-limit problem bites hardest when the PCC is close to the fault and the voltage collapses. If the residual PCC voltage is not too low, the inverter can still provide a meaningful, controlled fault-current contribution, set by its controls and its limits. “Limited fault current” does not mean “no contribution”.

Reduced and controlled fault current changes protection studies. Overcurrent relays, fuses, distance relays, directional elements and breaker-duty calculations may all be affected, because the current magnitude, angle, duration and sequence content can differ from a synchronous-machine network — sometimes enough to de-sensitise a scheme or require it to be redesigned.

Modelling implication. Do not carry synchronous-machine fault-current assumptions into an IBG protection study; represent the current limit, the active/reactive priority and the low-voltage cut-off, and check the result against the actual protection scheme.

Section 5

Difference 3 — Fast control response

Inverters can respond extremely fast, often within a fraction of a grid-frequency cycle, and because the behaviour is defined in software it is flexible. The inverter can be designed to act on a local voltage measurement, a local frequency measurement, a local fault, a transmission-system disturbance, a distribution-network requirement or an external command. That speed and flexibility are a genuine advantage.

But fast is not automatically good. If the controls are poorly designed, the inverter can over-react to small voltage or frequency variations and create abnormal system behaviour. This matters most in weak grids, islanded systems and systems with high inverter penetration. The fast response is therefore both an opportunity and a risk — which is precisely why it must be specified, tested and modelled rather than assumed.

Section 6

Difference 4 — The internal voltage source and operating modes

A synchronous generator has an internal induced voltage behind its reactance, normally higher than the terminal voltage and set by the excitation system. During a voltage sag, that internal voltage tends to drive more current into the grid, which helps support voltage and stability. A grid-following inverter has no such inherent source: it does not establish the grid voltage; it measures the existing voltage and injects current according to its control reference and current limit, usually around rated current or slightly above.

Grid-following versus grid-forming

Grid-following (GFL) — follows an existing grid voltage; normally uses a PLL or equivalent to synchronise; injects controlled active and reactive current; and can struggle in weak grids where the voltage angle is poorly defined.

Grid-forming (GFM) — establishes a voltage magnitude and frequency reference; behaves more like a controlled voltage source; can support weak-grid or islanded operation and energise a passive network; but is still limited by its converter current rating, its energy source and its protection.

Most IBGs today run in the P-Q (grid-following) mode; grid-forming is emerging and is key for high-IBG systems, though changing operating mode may require stopping the inverter first. One caution is worth stating plainly: grid-forming does not remove the need to model current limits, dc-side energy, protection and control tuning. A grid-forming inverter can improve system strength, but it is not an unlimited synchronous machine — push it past its current ceiling and it must protect itself like any other converter.

Table 2 — The two inverter operating modes at a glance.
Property	P-Q (grid-following)	U-F (grid-forming)
What it controls	Active and reactive power	Its own voltage and frequency
Grid angle	Follows the grid via the PLL	Sets the angle itself
Needs a stiff grid reference	Yes	No — can energise a passive network
Typical status	The common mode today	Emerging, key for high-IBG systems

Modelling implication. State the control mode explicitly; a grid-following model must include the PLL and current limit, and a grid-forming model must still include the current limit, the energy source and the protection — not an ideal voltage source.

Section 7

Difference 5 — Voltage support at transmission level

Large synchronous generators usually run in AVR mode — an automatic voltage regulator that regulates terminal voltage and supports voltage in the high-voltage transmission network, within the machine’s reactive capability. For large transmission-connected machines this is a major system-support function. Smaller machines may instead run in AQR mode (automatic reactive-power regulator) or APFR mode (automatic power-factor regulator), controlling reactive power or power factor rather than voltage directly, because they lack the reactive capability to hold voltage strongly.

Many IBGs have historically been operated at unity power factor: active power injected, reactive power approximately zero, and little or no voltage regulation. That may be acceptable for small distribution connections, but large transmission-connected IBGs are often required to provide voltage or reactive control at the PCC / POI. Large-scale IBG plants can support voltage — if they include oversized inverters, reactive capability, plant-level voltage control, or dedicated compensation such as a STATCOM (static synchronous compensator) or SVC (static var compensator). An IBG supports voltage only if its control mode and current rating allow reactive-current injection or absorption: the support is limited by the converter current circle, so if active power is high, less current capacity remains for reactive power unless the converter is oversized or active power is curtailed.

Modelling implication. Represent the reactive/voltage control mode at the PCC and the converter current circle (the active/reactive trade-off), not a fixed reactive capability borrowed from a machine.

Section 8

Difference 6 — Synchronising torque

This is one of the harder ideas, and one of the most important. Synchronising torque is the restoring torque that pulls a synchronous generator’s rotor back toward synchronism after a small angle disturbance — a physical consequence of the rotor angle and internal voltage interacting with the grid. When the rotor angle of one machine drifts away from the angle of the rest of the system, an electrical torque tends to pull it back. It is central to rotor-angle stability. The synchronising power coefficient between two machines \(i\) and \(j\) depends on their internal voltages, the reactance between them, and their angle difference:

\[ K_{ij} = \frac{\partial P}{\partial \delta_{ij}} = \frac{V_i\,V_j}{X}\,\cos\delta_{ij} \]

\(K_{ij}\): synchronising (power) coefficient between machines \(i\) and \(j\)
\(P\): active power transferred between the two machines
\(V_i,\ V_j\): internal induced (behind-reactance) voltages of the two machines
\(X\): reactance between the two internal voltages
\(\delta_{ij}\): angle difference between the two machines

In words: if the angle difference changes, the power transfer changes, and that change acts as a restoring effect that holds the machines in step. A larger \(\cos\delta_{ij}\) — angles not too far apart — means stronger synchronising torque; as \(\delta_{ij}\) approaches \(90^\circ\) it collapses toward zero.

Grid-following IBGs do not have classical synchronising torque, because they do not have a directly grid-coupled rotor angle. An IBG uses a PLL to measure the grid-voltage angle and inject current in the correct phasor direction — but that is not synchronising torque. A PLL helps the inverter follow the grid angle; it does not by itself create the physical restoring torque, which reacts to the angle difference between two machines — the restoring effect that holds them in step, positive while that difference stays below about \(90^\circ\) — not to the absolute angle the PLL measures. Future IBGs may be required to emulate something similar, but it is difficult: it can need wide-area angle measurements and communication infrastructure, and the reference angle to compare against changes when units trip or the network topology changes.

Section 9

Difference 7 — Loss of synchronism

Loss of synchronism is a synchronous-machine rotor-angle phenomenon. If rotor-angle stability is not maintained, the rotor angle runs too far from the rest of the system and cannot recover — the classical transient-stability problem. PV and full-converter plants have no physical rotor angle, so they do not lose synchronism in that classical sense. This is often misread as “IBGs have no stability issues” — which is wrong. An IBG still has to stay synchronised to the ac grid through its PLL, and it can still trip, lose PLL tracking, hit its current limit, enter momentary cessation, or become unstable through converter-control interaction. The careful statement is therefore: IBGs do not have classical transient rotor-angle instability, but they can still lose grid synchronisation or misbehave because of control and voltage-measurement problems.

Section 10

Difference 8 — Damping and power oscillation damping

Power systems can oscillate — local machine modes, inter-area modes, low-frequency electromechanical modes — and damping means reducing those oscillations by injecting or absorbing extra power in phase with the speed deviation, typically over a window of 5 to 10 seconds. A synchronous machine damps in two ways. Naturally, through its damper (amortisseur) windings: when the rotor speed deviates from synchronous speed, currents are induced in the windings and, by Lenz’s law, oppose the change. And with a power system stabiliser (PSS) added to the excitation system, which modulates the excitation to improve damping of rotor oscillations.

An IBG does not automatically provide the same damping, but its controls can be designed to. It can modulate active power, which damps oscillations directly, or reactive power, which damps them indirectly by moving voltage and therefore voltage-dependent loads. This is done with a power-oscillation-damping (POD) controller, and the same POD idea can sit on other power-electronic devices such as SVCs and HVDC (high-voltage direct-current) links.

IBGs can improve or worsen damping

IBGs may improve damping if their controls are designed and tuned for the relevant oscillation mode. They may worsen it if they displace synchronous machines that carried stabilisers, change power-flow patterns, reduce inertia, or introduce poorly damped converter-control modes of their own. Damping is therefore not a fixed property of the plant — it depends on tuning and on what the IBG displaced.

Section 11

Difference 9 — Frequency control

Frequency control has three broad layers — primary, secondary and tertiary. For synchronous plant it depends mainly on the prime mover and governor (steam, gas or hydro turbine); the generator itself is not the only factor. For an IBG, emulating frequency control means raising or lowering active power. Lowering is usually easy — an IBG can reduce active power quickly during over-frequency. Raising is the hard part, because many renewable sources already run at maximum available power — a PV plant in full sun cannot simply produce more unless it was previously curtailed. Upward response therefore requires headroom, storage, wind-rotor kinetic energy, or a dispatchable primary source.

There are two common solutions. Headroom means deliberately operating below available power so the plant can increase output when needed — a 100 MW-capable wind farm held at 90 MW keeps 10 MW of upward reserve. Curtailment creates that headroom, at the economic cost of spilling available renewable energy — a lost opportunity cost that, because the resource is variable, cannot be recovered. Storage is the other route: a battery energy storage system (BESS) can provide fast upward and downward active-power response, limited by its power rating, state of charge, energy capacity, ramp limits, thermal limits and reserve-restoration strategy. Keeping reserve has a cost for conventional plant too.

Over-frequency (LFSM-O)

Limited frequency-sensitive mode for over-frequency (LFSM-O) means reducing generation when the frequency rises above a threshold — an emergency corrective action that helps arrest an over-frequency. For many IBGs, reducing output is technically easier than increasing it: a PV inverter simply changes its reference, so over-frequency reduction is rarely a technical limitation. For synchronous plant the prime mover often limits how fast: a gas turbine cannot always cut fuel suddenly without risking a change in the air-fuel ratio and a flame-out, so it carries a rate limit. A complete model should include the start frequency, the droop / slope, the dead-band, the ramp limit, the minimum output, and the reconnection / restoration behaviour.

Modelling implication. Include the upward-response source (headroom, storage or rotor energy) and its limits when frequency support is studied; without an energy path, an IBG cannot sustain an upward response no matter what the control requests.

Section 12

Difference 10 — Fault ride-through capability

FRT = fault ride-through, the ability to remain connected during defined voltage disturbances. LVRT covers low-voltage events; HVRT covers high-voltage events. A ride-through curve states when the plant must remain connected, but the model must also represent what the plant does while connected: the current limit, the reactive-current support, the active-power reduction, the PLL behaviour, and any blocking, momentary cessation or trip logic. International Electrotechnical Commission (IEC) standards (IEC 60034-3) require a synchronous generator to withstand a terminal short circuit of any kind without failing, but the complete plant may still trip depending on the prime mover — some thermal plants include a shear pin in the shaft that can break during a severe voltage dip and trip the unit to protect mechanical equipment.

Historically, distributed IBGs often did not ride through severe three-phase faults, and it is worth separating two very different causes. One is a genuine inverter limitation: when the voltage becomes very low the inverter may not detect the voltage phase angle accurately. The other is an external contactor / control-supply limitation: the magnetic contactor between the inverter and the grid can lose coil excitation when the line voltage falls very low — below roughly 30% — and open, which is a contactor issue, not an inverter one. Modern techniques improve IBG ride-through: better frequency estimation, off-delay-release magnetic contactors, and an uninterruptible power supply (UPS) for the control and contactors. Large modern utility-scale wind plants now have strong fault ride-through; some turbines can ride through solid faults for up to about 3 seconds — which can be difficult for synchronous machines on mechanical and transient-stability grounds. Capability varies widely, so it is not safe to assume all IBGs ride through the same events: check the ride-through curve and internal protection against vendor data and the connection requirements.

Momentary cessation

Momentary cessation means the inverter temporarily stops, or greatly reduces, current injection while remaining connected. It is not a full trip, but it can remove active and reactive support during the exact period when the system needs it — and if many IBGs cease together, the recovery can be badly worsened. This is the behaviour explored in the large-voltage-deviation guide, and it must be modelled distinctly from a trip.

Modelling implication. Model both the ride-through envelope and the behaviour inside it — current limit, reactive priority, PLL, blocking, momentary cessation and trip — not the curve alone.

Section 13

Difference 11 — Reactive power and negative sequence

Reactive power has two distinct parts that are worth keeping separate: steady-state voltage–reactive (V-Q) control, and reactive current during network faults.

Normal operation. Synchronous generators usually have rated power factors of about 0.80 to 0.95, so reactive capability is designed in; they can generate or absorb reactive power within their capability curve. IBG reactive control at the PCC / POI comes in familiar modes — voltage regulation, a fixed reactive power, a fixed power factor, or a Q(V) droop — usually coordinated by a plant-level controller. Many distributed IBGs historically ran at unity power factor, injecting no reactive power; an inverter already at full active power, and not oversized, may have no spare current for reactive power, so providing reactive power at full output can require oversizing.

During a fault. A synchronous machine increases reactive current immediately, from machine physics and its internal voltage, which helps support voltage during the disturbance. An IBG instead provides dynamic reactive-current injection or absorption, bounded by the current limit and the active/reactive priority, shaped by the LVRT/HVRT logic, and delayed by measurement and control compared with the machine’s instantaneous physics.

“An IBG cannot sense voltage instantaneously”

The converter measures voltage through sensors, filtering and digital control. This response can be very fast, but it is still not the same as the electromagnetic response of a synchronous machine, whose reactive current rises the instant the flux changes. Where the timing matters, the model should include a measurement delay, a control delay, or validated response behaviour — not an idealised instantaneous reaction.

Negative-sequence current

Negative-sequence current appears during unbalanced faults. A synchronous machine naturally allows it to flow — although it can heat the rotor and is limited by protection — whereas many IBGs are deliberately designed to suppress it unless the grid code requires otherwise. Controlled negative-sequence injection can help support the depressed phases during an unbalanced fault, but it must be represented correctly in an EMT or sequence-capable model. As an example of grid-code evolution, VDE-AR-N 4120 — a German high-voltage connection rule — includes requirements for converter behaviour during unbalanced faults, including negative-sequence current capability. Do not imply that this exact requirement applies everywhere; treat it as one example of where the rules are heading.

Modelling implication. Represent the steady-state V-Q mode and, separately, the fault-time reactive-current behaviour with its current limit and priority; and if unbalanced faults matter, use a sequence-capable or EMT model rather than a positive-sequence RMS one.

Section 14

Difference 12 — Harmonics

Harmonics are voltage or current components at integer multiples of the fundamental frequency — for a 50 Hz system the 5th harmonic is 250 Hz and the 7th is 350 Hz. Emission. Harmonic currents from a synchronous generator are usually small — air-gap flux harmonics, slot harmonics and construction effects — and generally negligible next to inverter concerns. Inverters, by contrast, can produce non-sinusoidal currents because of switching, controls, and interaction with filters and grid impedance; the emission depends on the inverter technology, the control strategy, the converter design, the transformer coupling and the harmonic voltages already present in the ac grid. Emissions must therefore be assessed before connection, against IEC standards and national limits.

Harmonic-voltage reduction. This is a subtler point. For low-order harmonics, a synchronous machine presents a low impedance — roughly its small subtransient reactance — giving harmonic currents a low-impedance path and helping to reduce harmonic voltages in the network. A voltage-source converter (VSC), however, presents a harmonic impedance that is control-dependent, not a simple passive reactance: it can absorb harmonics by behaving as an impedance through its control, or it can amplify them if the control interacts badly with the network. The consequence for modelling is that you cannot treat an inverter as a passive machine.

Modelling implication. RMS models are not suitable for harmonic emission or waveform-distortion studies; use EMT, harmonic-domain tools, frequency scans, impedance models or a validated vendor harmonic model, chosen for the study objective.

Section 15

Difference 13 — Black start

Black start is the ability to energise part of the power system without an external grid supply, to help restart the system after a full or partial blackout. Most large stations need an external electrical supply to start their auxiliary systems; a black-start-capable synchronous unit can instead start from its own auxiliary supply and establish voltage and frequency through its generator and prime mover, energising part of the network and supplying the inrush currents of restoration. An IBG can support black start only if it has grid-forming capability, a suitable energy source, control power, protection coordination, transformer-energisation capability and enough fault-current / overload capability for the restoration steps.

Why black start is hard for IBGs

Black start with IBGs requires: grid-forming control; a sufficient energy source or storage; a stable voltage / frequency reference; the ability to energise transformers and cables; management of inrush current with limited overload; protection coordination against a low fault-current contribution; a load-pickup strategy; and coordination with other grid-forming and grid-following resources. It is not impossible — grid-forming inverters are changing this — but it is more limited than a traditional synchronous black-start provider, and it must be modelled with a restoration-specific, grid-forming model.

Section 16

Difference 14 — Not all IBGs are the same

“IBG” is a family, not a single device. PV, wind and battery storage may all connect through inverters, but their primary energy constraints differ, and that shows up in frequency response, recovery and long-duration support even when the converter terminals look similar. PV depends on irradiance and has no upward reserve at full sun unless curtailed. Wind depends on aerodynamic power and rotor speed, and can lend short-term rotor kinetic energy. A BESS depends on its state of charge and its power / energy rating, and can move power both ways quickly for a bounded time. So two plants that look identical electrically at the converter can behave very differently in a frequency event.

Wind itself splits by topology, and the two main types must not be treated as identical in fault response, inertia emulation or control-interaction studies. A Type-3 (DFIG) wind-turbine generator has its stator directly connected to the grid with a rotor-side converter, so it is not fully decoupled and its fault contribution and dynamics sit between a synchronous machine and a full converter. A Type-4 (full-converter) WTG connects the generator to the grid entirely through the converter, so it behaves like a controlled current (or, if grid-forming, voltage) source. PV and battery plants are full-converter-connected and resemble Type-4 electrically.

A minor practical difference rounds out the list: synchronous machines are large rotating machines and usually need more periodic maintenance and longer downtime, while an inverter carries a lighter mechanical-maintenance burden — a real, if secondary, advantage. Modelling implication. Fix the plant type (PV, BESS, Type-3 or Type-4 wind) before choosing a model; a generic “IBG” block is only as good as the technology assumptions baked into it.

Section 17

Why it matters — and what IBGs do well

As systems move from fossil-fuelled synchronous plant to inverter-based generation, they can lose several services that used to come for free: inertia, high fault current, voltage support, synchronising torque, damping, black-start capability and natural fault behaviour. To compensate, grid codes increasingly require IBGs to provide advanced functions — frequency regulation, reactive and voltage control, low-voltage ride-through, power-oscillation damping, synthetic inertia or fast frequency response, fault-current injection and negative-sequence support. But none of these is a natural feature of an IBG; each depends on control design, inverter size, energy availability and the grid code.

IBGs are different, not worse

With the right controls, IBGs offer real strengths: very fast, flexible control; frequency and reactive/voltage regulation; low-voltage ride-through; independent active and reactive current control; immunity to some mechanical torque disturbances; quick output reduction on over-frequency; and lighter mechanical maintenance. In some respects they out-perform synchronous machines. The only catch is that the behaviour must be specified, controlled, tested and correctly modelled — never assumed.

Section 18

Modelling implications for dynamic studies

This is the practical conclusion. For synchronous machines, most behaviours are physically inherent and fairly standard. For IBGs the behaviour is not universal — it lives in the controller — so an IBG model must include the specific capabilities the study needs. In an EMTP® or other EMT study, that means matching the model to the question being asked:

Frequency stability — represent the frequency response or synthetic inertia, if present, with its headroom / storage.
Voltage stability — represent the reactive-power and voltage control at the PCC, with the current limit.
Fault ride-through — represent the current limits, fault logic, voltage measurement, PLL behaviour, momentary cessation and ride-through settings.
Weak-grid behaviour — represent the PLL and control interactions, since these dominate at low short-circuit ratio.
Harmonics — an RMS model is not enough; treat the converter’s harmonic behaviour separately.

RMS or EMT?

Use RMS when the study is dominated by fundamental-frequency active/reactive-power behaviour — bulk frequency stability, long-term voltage stability, or large-system screening with validated IBG models. Use EMT when the result depends on waveform detail, weak-grid converter interaction, PLL / current-control dynamics, unbalanced faults, negative-sequence behaviour, harmonics, switching, momentary cessation or vendor-specific protection.

Do not assume an IBG…

…provides inertia — unless it is enabled and energy is available;
…provides high fault current — it is current-limited;
…behaves as a voltage source — not if it is grid-following;
…supports voltage at full active output — not without current headroom;
…injects negative-sequence current — not unless designed or required to;
…rides through all faults — only within its FRT curve and internal protection;
…can black-start — only with grid-forming and proven energisation capability.

Practice bears this out. A CIGRE survey of system operators found the simple ‘negative-load’ representation still the most common IBG model for frequency- and rotor-angle-stability studies, with RMS models favoured for those and EMT models used for short-term voltage stability, fault ride-through and other fast phenomena — the model follows the study, not the other way round. Some capabilities are available commercially but are not yet standardised in algorithm, performance requirement, implementation, compliance assessment or generic modelling, so it is difficult to build a single generic IBG model valid for every study — the model has to be chosen and configured for the job in hand, and correctly initialised before the disturbance is applied.

Section 19

Weak grids, SCR and the area of validity

IBG behaviour becomes far more sensitive in weak grids, because a change in converter current moves the local voltage magnitude and angle more strongly. Grid-following PLLs, current controllers and plant voltage controllers can then interact with the network impedance, so the grid strength and any nearby converters matter. Grid strength is screened with the short-circuit ratio (SCR) — commonly the grid short-circuit power at the connection point divided by the plant rating. A high SCR indicates a strong grid; a low SCR indicates a weak grid. SCR is a screening measure, not a universal stability guarantee: two plants at the same SCR can behave differently depending on their controls.

Every IBG model has an area of validity — the conditions under which it is reliable. An RMS model, for instance, may represent voltage or frequency support correctly only when the grid is not too weak, the SCR is above a minimum value, the voltage stays within a defined range, the controller remains in its normal operating mode, the inverter has not reached its current limit, and the PLL remains stable. Outside that envelope the same model can quietly give the wrong answer.

Before modelling an IBG, confirm…

plant type: PV, BESS, Type-3 wind, Type-4 wind or other;
grid-following or grid-forming mode;
converter current rating and overload capability;
active / reactive current priority;
voltage / reactive control mode;
FRT curve and internal protection;
momentary-cessation settings;
frequency response and headroom;
ROCOF / vector-jump / anti-islanding settings;
negative-sequence capability;
harmonic model or impedance data;
SCR / grid strength;
the model validity range and the software version.

And one warning that underlies all of it: a generic IBG model is not automatically valid because it runs. Validate it against grid-code requirements, vendor documentation, commissioning tests, measured disturbance records, or an EMT benchmark where the behaviour is critical. When in doubt, an EMT representation that keeps the real control structure is safer than an RMS shortcut — most of all for weak grids, low short-circuit levels, severe faults and high inverter penetration, exactly the cases where an out-of-validity model looks plausible but is wrong.

Common mistakes

The misconceptions that most often distort an IBG study — each the flip side of a difference above:

Ten traps to avoid

Treating an IBG as a synchronous generator with a different rating.
Assuming synthetic inertia is the same as physical inertia.
Ignoring headroom or storage for upward frequency response.
Assuming IBG fault current is high enough for existing protection.
Modelling LVRT as only a curve and ignoring current limits and protection.
Assuming grid-forming has unlimited fault current.
Ignoring negative-sequence behaviour in unbalanced faults.
Using RMS for harmonic or switching studies.
Assuming all PV, BESS and wind inverters behave the same.
Ignoring vendor-specific controls and protection.

Section 20

The differences at a glance

A fuller comparison, mapped to the stability phenomenon each characteristic affects and split by whether the IBG has only minimum functions or the advanced capabilities that evolving grid codes now require. The pattern is consistent: what a synchronous machine gives naturally, a minimum-function IBG lacks, and an advanced IBG can provide only through control — often with conditions attached.

Table 3 — Synchronous generator versus IBG, before and after grid codes evolve, mapped to the stability phenomenon each characteristic affects (after the CIGRE brochure’s Table 2.1).
Characteristic	Synchronous generator	IBG — minimum functions	IBG — advanced (after grid codes)
Inertia / rotating mass Frequency stability	Yes	No	Yes, prime-mover dependent; needs headroom or storage; not an exact emulation
Frequency response (primary–tertiary) Frequency stability	Yes	No	Yes, prime-mover dependent; needs headroom
Limited frequency sensitive mode Over-frequency	Yes	No	Yes, prime-mover dependent
Constant (internal) voltage source Voltage stability	Yes, internal induced voltage	No, when grid-connected	Yes, but needs stiff U-F (grid-forming) operation; may need oversizing
Transmission-level voltage support Voltage stability	Yes, large machines with AVR	No	Yes, large-scale IBG only, often with SVC / shunt compensation
Reactive power, V-Q steady state Voltage stability	Yes, per PQ capability	No	Yes, larger inverter or reduced active power
Reactive current during faults Voltage / rotor-angle stability	Yes, immediate	No	Yes, with some delay, within rated current
Synchronising-torque capability Rotor-angle stability	Yes	No	Yes but almost infeasible; needs delay-free wide-area angle measurement
Damping torque / POD Rotor-angle stability	Yes, damper windings + PSS	No	Yes, POD functionality
Loss of synchronism Rotor-angle stability	Yes	Not applicable to IBG	—
Fault ride-through Transient / frequency stability	Yes	No	Yes, prime-mover dependent
Short-circuit contribution Protection	Yes	No	Yes, limited to \(\approx 1.1\) pu unless the short-term rating exceeds it
Harmonic emission Power quality	Low	Yes, including the power-line-communication (PLC) band	Assessed before connection
Harmonic-voltage reduction Power quality	Yes, low-order (subtransient)	Yes	Yes
Control-response capability Voltage & frequency stability	Fast, time-constant limited	Inverter very fast; limited by measurement / prime-mover delay	—
Overload capability (a few seconds) Miscellaneous	Yes	Limited, near-negligible (device-dependent)	Yes, with significant oversizing
Maintenance Miscellaneous	Regular	Lower for the inverter; prime-mover dependent	—

Two reading notes: “prime-mover dependent” means the advanced capability exists only if the energy source behind the inverter can actually supply it (headroom, ramp rate, stored energy); and a dash means the row adds no distinct advanced case beyond the minimum-function column. The middle column is the IBG “as installed” in much of the existing fleet; the right-hand column is what modern grid codes now ask new plant to provide.

Key points

Physics for one, control for the other

A synchronous generator provides many grid services naturally, through physics.
An IBG provides comparable services only through controls, current rating, energy availability and grid-code requirements.
Grid-following IBGs normally behave as controlled current sources, not internal voltage sources.
Grid-forming IBGs can establish voltage and frequency but still have converter limits.
IBG fault current is limited and changes protection assumptions.
Frequency support requires headroom, storage or controlled recovery.
Voltage support is limited by current rating and P/Q priority.
FRT needs both a ride-through envelope and correct behaviour inside it.
Harmonics and negative-sequence behaviour are converter- and control-specific.
Model selection must follow the phenomenon: RMS for bulk phasor dynamics, EMT or specialised models for waveform, weak-grid, control, sequence and harmonic behaviour.

High IBG penetration changes the fundamental dynamic behaviour of the power system: the missing natural features of synchronous generation can often be replaced by advanced inverter controls, but they are not automatically present, and they are not always fully equivalent. For the control layers behind these behaviours, see the full-converter control, plant-controller and converter-protection guides.

References

This page draws on the CIGRE/CIRED technical brochure on inverter-based generation; the EMTP® documentation informs the modelling framing; the German high-voltage connection rules are the reference for negative-sequence injection; and the Wiley–IEEE reference work provides the converter-control background.

CIGRE/CIRED Joint Working Group, Modelling and Dynamic Performance of Inverter-Based Generation in Power System Transmission and Distribution Studies. CIGRE Technical Brochure.
EMTP®, Renewable and Inverter-Based Generation Modelling — Documentation and Application Notes. Powersys / EMTP®.
VDE-AR-N 4120, Technical Requirements for the Connection and Operation of Customer Installations to the High-Voltage Network (TAR High Voltage). VDE-FNN, Germany.
R. Teodorescu, M. Liserre and P. Rodríguez, Grid Converters for Photovoltaic and Wind Power Systems. Chichester, UK: Wiley–IEEE Press, 2011.

Part 1 Reading now

Characteristics of Inverter-Based Generation

What sets inverter-based generation apart from synchronous machines, and why that difference drives every modelling choice that follows.

Series progress 1 of 12

Part 2

Inverter Characteristics for IBG Modelling

Inside the inverter: topology, the dc/ac chain, control, protection and the capability functions grid codes require.