Renewable Modelling · GSC Fault Behaviour

GSC Fault Behaviour in EMTP®

Fault-ride-through current priority and decoupled sequence control

During grid faults, the grid-side converter (GSC) must decide how its limited current capability is shared between active-power transfer, reactive-current support and sequence-control requirements. The converter has a single current circle, and two mechanisms govern how that finite current is allocated: fault-ride-through (FRT) current priority, which decides whether active or reactive current wins when the converter is current-limited; and decoupled sequence control (DSC), which handles the second-harmonic oscillations an unbalanced fault stirs up. FRT decides which current gets priority; DSC decides how the sequence components are controlled — and both compete for the same finite current.

Reading time ≈ 20 min · FRT priority & decoupled sequence control

During a grid fault the grid-side converter (GSC) has a finite current capability, and that single limit governs everything that follows. Two distinct but related mechanisms decide how the current is used. Fault-ride-through (FRT) current priority — FRT being the control behaviour that keeps the converter connected and supports the grid through specified voltage disturbances — decides whether active or reactive current wins when the converter is current-limited. Decoupled sequence control (DSC) handles the second-harmonic oscillations that an unbalanced fault stirs up. The cleanest way to hold them apart: FRT decides which current gets priority; DSC decides how the sequence components are controlled. Both run into the same hard limit — the converter’s finite current.

Abbreviations used on this page

GSCGrid-side converter

FRTFault ride-through

DSCDecoupled sequence control

FSCFull-scale converter

\(V^{+}\)Positive-sequence voltage

\(I_g^{\lim}\)Total converter (GSC) current limit

d–qDirect / quadrature rotating frame

PIProportional–integral controller

\(2\omega\)Twice fundamental frequency (second harmonic)

PCCPoint of common coupling

WTCWind-turbine controller

EMTP®Electromagnetic Transients Program

Key idea

The GSC has a maximum current circle (\(i_{dg}^{2}+i_{qg}^{2}\le (I_g^{\lim})^{2}\)), so it cannot maximise active and reactive current at once — a priority rule is needed.
Normal operation: active current first (the d-axis holds the dc-link and exports power), reactive gets the remaining margin. During FRT (deep voltage dip) the priority reverses — reactive current first for voltage support, active gets the leftover. FRT activates when \(|1-V^{+}|>V_{\text{FRT-ON}}\) and releases when it falls below \(V_{\text{FRT-OFF}}\) for \(t_{FRT}\); the limits themselves may also change.
Under unbalance the terminal voltage gains a negative-sequence component, whose cross-terms with the positive sequence produce second-harmonic (\(2\omega\)) power oscillations that ripple the dc-bus voltage. Decoupled sequence control splits the control into separate positive- and negative-sequence dq loops to shape this.
Setting the negative-sequence reference to zero suppresses the ripple; setting it per grid code contributes negative-sequence current. DSC’s effectiveness is limited by the converter rating, the FRT positive-sequence reactive demand, the fault type and the electrical distance (a closer fault needs more current).

Key terms used on this page

01Fault ride-through

Staying connected and supporting voltage through a grid fault, as grid codes require.

02Current circle

The converter’s capability limit: \(i_d^{2}+i_q^{2}\) cannot exceed the squared maximum current.

03Current priority

The rule deciding which of active or reactive current is served first within the current circle.

04FRT thresholds

\(V_{\text{FRT-ON}}\) / \(V_{\text{FRT-OFF}}\): the voltage deviations that enter and leave FRT mode.

05Release time

\(t_{FRT}\): the delay before leaving FRT once the voltage has recovered, to avoid chattering.

06Negative sequence

The reverse-rotating symmetrical component that appears in the voltage under unbalance.

07Second-harmonic ripple

The \(2\omega\) oscillation in power and dc-bus voltage caused by positive–negative sequence cross-terms.

08Decoupled sequence control

Controlling positive- and negative-sequence currents in separate dq loops to shape the unbalance response.

09Sequence loops

The two dq controllers — one for positive sequence, one for negative — whose references combine.

10Negative-seq reference

The commanded negative-sequence current: zero to suppress ripple, or a grid-code value to contribute.

11Converter rating limit

The finite current that all competing objectives — FRT, DSC, active transfer — must share.

12Electrical distance

How close the fault is electrically; a closer unbalanced fault deepens the asymmetry and demands more current.

Section 1

What the GSC does when the grid goes abnormal

Balanced, steady operation is the easy case. The interesting — and grid-code-critical — behaviour is what the converter does during a fault or an unbalance. Two questions arise, and they have separate answers. When the converter runs out of current capability, which current does it serve first? And when the grid is unbalanced, how does it stop the resulting oscillations from upsetting the dc link? FRT answers the first; DSC answers the second. They are not separate controllers bolted on outside the GSC — they are special modes of the same current-reference and current-limiting structure.

Two answers

FRT = “which current gets priority when the converter is stressed?” DSC = “how do we control positive- and negative-sequence currents separately so the second-harmonic oscillations are reduced?”

Section 2

The converter current circle

Everything in FRT follows from one hard fact: the converter has a maximum current. The current circle is the converter’s maximum current capability in the d–q plane — increasing one current component reduces the margin available for the other. The d- and q-axis currents share that single budget, so they lie inside a current circle:

\[ i_{dg}^{2} + i_{qg}^{2} \le \big(I_g^{\lim}\big)^{2} \]

\(i_{dg},\ i_{qg}\): d-axis (active) and q-axis (reactive) GSC currents
\(I_g^{\lim}\): limit on the total GSC current

Because the active and reactive currents cannot both be maximised at once, a priority rule must decide which gets the current budget first.

Section 3

Normal operation: active current first

In normal, balanced operation the grid-side converter gives priority to active current, because it must regulate the dc-link voltage and export the available power. The d-axis current holds the dc-link and delivers that power, so it is limited first; the q-axis (reactive) current then takes whatever margin remains from the total current budget:

\[ i_{dg}' < I_{dg}^{\lim}, \qquad i_{qg}' < I_{qg}^{\lim} = \sqrt{\big(I_g^{\lim}\big)^{2} - \big(i_{dg}'\big)^{2}} \]

\(I_{dg}^{\lim}\): limit on the d-axis (active) current
\(I_{qg}^{\lim}\): limit on the q-axis (reactive) current — the remaining margin once \(i_{dg}'\) is chosen
\(I_g^{\lim}\): limit on the total GSC current

There is no emergency, so the controller keeps exporting power: active current is served first, and reactive support uses the leftover capacity.

Doing the opposite in normal operation would needlessly sacrifice active-power transfer for reactive support the grid is not asking for — so active-first is the right economic and operational choice while conditions are normal.

Section 4

During FRT: reactive current first

When a deep voltage dip is detected, the situation changes. The grid no longer mainly needs active-power export; it needs voltage support, and reactive current supports voltage far more directly than active current in that short fault window. Many grid-code requirements therefore prioritise positive-sequence reactive current during deep voltage dips, so the GSC reverses the priority: reactive current is now served first, and active current is reduced to whatever remains within the current limit. The same current circle still applies — only the allocation rule changes, by reversing which axis is limited first.

Table 1 — Current-priority allocation in normal operation and during fault ride-through.
Mode	First Priority	Gets the Remaining Margin	Reason
Normal	Active current \(i_{dg}\) (dc-link / power transfer)	Reactive current \(i_{qg}\)	No emergency — keep exporting power
FRT (deep dip)	Reactive current \(i_{qg}\) (voltage support)	Active current \(i_{dg}\)	Voltage support prioritised — active current sacrificed first

So during ride-through the controller reduces the active-current demand first, freeing converter current capacity for reactive-current injection. For that moment the reactive support outweighs the normal active-power transfer — though the exact reactive-current obligation varies by grid code and project.

Section 5

When FRT is activated and released

FRT mode is entered and left on the positive-sequence voltage deviation. It is activated when the deviation exceeds an upper threshold, and released only when it has fallen below a lower threshold and stayed there for a release time:

\[ |1 - V^{+}| > V_{\text{FRT-ON}} \;\Rightarrow\; \text{enter FRT}; \qquad |1 - V^{+}| < V_{\text{FRT-OFF}}\ \text{for}\ t_{FRT} \;\Rightarrow\; \text{leave FRT} \]

\(|1 - V^{+}|\): positive-sequence voltage deviation from nominal (1 pu)
\(V_{\text{FRT-ON}}\): activation threshold (enter ride-through)
\(V_{\text{FRT-OFF}}\): release threshold (lower than activation)
\(t_{FRT}\): release time the deviation must stay below \(V_{\text{FRT-OFF}}\)

The lower release threshold and the release time together act as hysteresis plus a delay: the controller does not snap back at the first flicker of recovery, which prevents chattering between normal and FRT modes and gives a clean transition.

Section 6

The limits themselves may change

FRT is not only a change of priority. The limits for the d-axis, q-axis and total GSC current may also be different during a fault. Some models and grid-code implementations allow a different overload capability for short disturbances, special fault-current limits, particular positive- / negative-sequence priorities, and different thermal allowances for the brief fault duration — but any such changed limits or temporary overload capability must be confirmed against the OEM or project data. So FRT can shift the whole allowable operating envelope, not just the order in which the budget is filled; the full detail depends on the model implementation and the grid code.

A control action, not a trip

FRT current priority is a control action taken while the converter stays connected and keeps injecting current. It is distinct from the protection that blocks or trips the converter when a limit is genuinely exceeded — priority reallocates the available current; protection removes the converter from service.

Section 7

Unbalanced grids and negative sequence

Balanced-fault FRT priority is only half the story. For unbalanced faults, current priority alone is not enough, because the negative-sequence voltage that appears at the terminal creates second-harmonic power and dc-link oscillations that a priority rule cannot address. The neat thing about dq control is that balanced three-phase sinusoids become constants, and active and reactive can be decoupled — but that is only perfectly true when the system is balanced. Under unbalanced loading or an unbalanced fault, the terminal voltage of a full-converter turbine contains a negative-sequence component as well as the positive sequence. A single synchronous dq frame no longer sees purely constant quantities: the opposite-sequence component appears as an oscillatory term in that rotating frame, at twice the fundamental frequency.

Section 8

Why negative sequence causes second-harmonic ripple

When positive- and negative-sequence components coexist, the instantaneous power picks up oscillatory cross-terms at twice the fundamental frequency, so the converter power is no longer steady:

The mechanism is a short chain: unbalanced voltage → negative sequence → interaction with the positive sequence → \(2\omega\) power oscillation → dc-link ripple → disturbance to the GSC control and PLL.

\[ p(t) = P_0 + P_{2}\cos(2\omega t + \phi) \]

\(p(t)\): instantaneous GSC power
\(P_0\): steady (average) power
\(P_{2}\): amplitude of the \(2\omega\) pulsation from the positive–negative sequence cross-terms
\(2\omega\): twice the fundamental angular frequency (100 Hz at 50 Hz, 120 Hz at 60 Hz)

The \(2\omega\) power pulsation upsets the dc-link power balance, so the dc-bus voltage oscillates — and the PLL and PI loops dislike that. The problem is not that negative sequence is “bad”; it is that it stresses the dc link and the control system.

Section 9

Decoupled sequence control

There are two broad ways to handle this. One is to tolerate the ripple and filter it — keep a simpler control and accept the second harmonic. The other, more capable, is decoupled sequence control (DSC): separate the control into a positive-sequence channel and a negative-sequence channel, and assign references to each. Instead of one dq controller wrestling with a waveform that contains mixed sequence content, the model runs two dq loops — one for the normal balanced power transfer (positive sequence) and one that explicitly controls the negative-sequence current — and builds the converter reference from the combination. That is the “sequence decoupling” the name refers to, and it lets the converter directly shape the sequence behaviour and suppress the unwanted oscillatory terms.

It helps to see why two dq loops are enough to cancel the pulsation. The single amplitude–phase term written above unpacks into two independent parts — a cosine and a sine coefficient at \(2\omega\):

\[ p(t) = P_0 + P_{c2}\cos(2\omega t) + P_{s2}\sin(2\omega t) \]

\(P_{c2},\,P_{s2}\): the two independent \(2\omega\) power coefficients set by the positive–negative sequence cross-terms
\(i_{d}^{+},\,i_{q}^{+}\): positive-sequence d- and q-axis current references
\(i_{d}^{-},\,i_{q}^{-}\): negative-sequence d- and q-axis current references

DSC controls four current references — the two positive-sequence and two negative-sequence dq quantities — so it has the freedom needed to drive both \(P_{c2}=0\) and \(P_{s2}=0\) and so reduce the \(2\omega\) ripple on the dc bus, subject to the available current. The positive-sequence reactive current is still pinned by the FRT demand, so only the reserve left over from that requirement is available for the negative-sequence cancellation — which is why the suppression weakens as the fault moves electrically closer and the demanded current approaches the GSC rating. The full solver — the four current loops and the sequence frames — is set out in the decoupled sequence control implementation; here the focus is fault behaviour and priority.

Section 10

Setting the negative-sequence reference

The power of DSC is in what you command for the negative-sequence current. Two practical choices stand out:

Table 2 — Setting the negative-sequence current reference in decoupled sequence control.
Negative-Sequence Reference	Effect
Set to zero	Suppresses negative-sequence current injection, reducing the second-harmonic ripple in power and dc-bus voltage — the simplest form of DSC.
Set per grid code	Contributes the required negative-sequence current during the unbalanced fault, as some grid codes demand.

So DSC is not only a way to cancel negative sequence — it is a framework for controlling the sequence components separately. A zero reference suppresses negative-sequence current injection and protects the dc side and the control from the oscillations; a non-zero reference allows specified negative-sequence support where a grid code requires it. That flexibility is what makes the method powerful.

Section 11

Why DSC competes with FRT

Here is the central engineering trade-off. The effectiveness of DSC is limited by the GSC current rating, the FRT requirement set by the grid code, the fault type and the electrical distance to the fault. The reason is simply that, during a fault, the converter current capability is finite, and several goals compete for it at once: positive-sequence reactive current for FRT, negative-sequence current control for oscillation suppression, dc-link stabilisation, and as much active-power transfer as possible. One converter has only so much current, so spending capability on sequence decoupling may reduce something else. During a severe fault the positive-sequence reactive-current support may consume most of the converter current margin, leaving limited capacity for negative-sequence control — so the freedom for DSC shrinks exactly when the fault is worst.

Electrical distance matters because a closer unbalanced fault causes deeper voltage asymmetry at the turbine terminal, stronger negative-sequence effects and more severe oscillatory terms — so more converter current is needed to cancel them. A more distant fault produces milder unbalance, which DSC can clean up more easily, sometimes without sacrificing active power; a near fault may force a compromise.

Section 12

FRT and DSC together

Putting the two together: during an unbalanced fault the GSC must do both jobs at once — satisfy the FRT reactive-current requirement on the positive sequence, and manage the positive- and negative-sequence currents properly through DSC. Both draw on the same finite converter current, which is exactly why the current limiter is dynamic: it arbitrates between these competing demands according to the fault and the priority rules. FRT decides the priority; DSC decides the sequence control; and the converter rating sets the ceiling on what either can achieve.

One converter, several objectives

The real difficulty under an unbalanced fault is not any single mechanism but their collision on a single current budget — FRT priority, sequence decoupling, dc-link support and active transfer all competing at once.

Section 13

Key points

The GSC has a finite current capability — a single current circle that every demand must share.
Normal operation prioritises active current, holding the dc-link and exporting power; reactive current takes the remaining margin.
Deep voltage dips usually prioritise positive-sequence reactive current for voltage support, with active current reduced to whatever remains within the current limit.
Unbalanced faults create \(2\omega\) oscillations in power and dc-bus voltage, driven by the negative-sequence terminal voltage.
DSC can reduce those negative-sequence effects, but only within the current margin left after the FRT reactive-current demand, and subject to fault severity and asymmetry.

References

EMTP® Documentation and Application Notes. Powersys / EMTP®.

Built on EMTP® · Expert spotlight

Henry Gras

Chief Operating Officer, PGSTech · Montréal, Canada

Henry Gras delivers the EMTP® University course “EMT Simulation and Analysis of Large-Scale Power Systems with Renewables” and works daily with the tool this article is written around.

Henry is based in Montréal, where he is Chief Operating Officer of PGSTech, the company responsible for EMTP® engineering services, commercialisation and continuing software development. He holds a master’s degree from Polytechnique Montréal, where he worked on electrical-machine research, and previously completed an engineering degree at École Centrale de Lyon in France.

Readers who want a structured programme on EMT simulation of large-scale power systems with renewables will find his EMTP® University course an excellent next step.

Henry’s technical expertise covers electromagnetic transient simulation, renewable-energy integration, power-system modelling, electrical machines, protection and specialist transient studies including TRV, transformer energisation, ferroresonance, insulation coordination and power quality.

Part 12 Reading now

GSC Fault Behaviour: FRT and Decoupled Sequence Control

Why fault-ride-through needs current priority, and why unbalance calls for decoupled sequence control.

Series progress 12 of 30

Part 13

Decoupled Sequence Control Implementation

Inside the controller — sequence separation, the four current references, the decoupled solver, current limiting and FRT priority.