VSC-HVDC Outer-Loop Control: DC Voltage, P/Q Control & Grid Stability

Part A — VSC-HVDC Outer-Loop Control at a Glance

A voltage-source converter (VSC) HVDC terminal is a power electronic machine whose behavior is defined almost entirely by its controls. Its inner current loop decides how fast and how precisely current is delivered. Its outer loops decide what that current is for: holding active power, holding reactive power, regulating DC voltage, regulating AC voltage or supporting frequency. The same hardware can stabilize a weak grid or destabilize it, damp an inter-area oscillation or ignore it, ride through a fault or trip on DC overvoltage — depending on which outer loop is selected and how it is coordinated with the other terminal.

The outer-loop menu

Defining an equivalent DC inertia constant per terminal:

the per-unit dynamics become:

MMC converters typically store on the order of 30 to 40 kJ per MVA, so Hdc is about 0.03 to 0.04 s per terminal.

3.2 Worked example: how fast DC voltage rises

With Hdc = 0.04 s and a 1.0 pu power imbalance, the time for DC voltage to rise from 1.0 to 1.1 pu is:

The DC voltage does not oscillate; it drifts. In point-to-point links, one terminal (usually the inverter, but sometimes the rectifier) controls DC voltage and the other controls active power. In multi-terminal systems, several terminals share DC voltage control through DC-voltage droop:

so that power imbalances are shared among terminals in proportion to their droop gains.



3.4 Control modes for a point-to-point link

3.5 DC faults and their effect on control strategy

A DC-side fault — a cable insulation failure or, for overhead DC lines, a lightning-induced pole-to-ground flashover — collapses DC voltage. Half-bridge MMCs cannot block DC fault current, because their freewheeling diodes form an uncontrolled rectifier from the AC side; the fault must be cleared by opening AC breakers, which takes several cycles and interrupts transfer. Full-bridge or hybrid submodules can drive DC current to zero by inserting negative voltage, and HVDC circuit breakers allow selective DC fault clearing in multi-terminal grids.

Outer-loop design must include post-fault restart sequences: which terminal re-establishes DC voltage first, how fast power ramps back, and how the AC systems experience the temporary loss of transfer.

4. Weak AC systems

4.1 Short-circuit ratio for a VSC terminal

VSC terminals can operate at much lower SCR than LCC terminals, but low SCR still limits power transfer, raises voltage sensitivity and makes control interactions more likely.

4.2 Static voltage at a weak bus



For a converter injecting P and Q into a Thevenin source E behind reactance X, the bus voltage V satisfies:

4.3 Worked example: Q control versus voltage control at SCR 1.8



With SCR = 1.8 (X = 0.556 pu), E = 1.0 pu and P = 0.8 pu:

At unity power factor, the bus sits at 0.85 pu — below the typical 0.9 pu lower limit for normal operation — and the converter's voltage sensitivity is high. AC voltage control supplies about 0.19 pu reactive power and holds 1.0 pu. In Keentel's fault simulation, a remote fault that drops the source to 0.35 pu for 150 ms collapses the bus voltage under fixed Q (the operating point has no solution once power cannot be delivered), while AC voltage control supports about 0.82–0.85 pu during the fault and returns to 1.0 pu with a brief 1.07 pu overshoot.

4.4 Worked example: SCR across operating conditions

A 1,000 MW receiving terminal at a 345 kV substation sees different short-circuit levels as the network changes:

A control philosophy tuned only for the all-in-service case can fail under credible outages. Studies must cover the full SCR range expected over the project's life, including future retirements of synchronous generation, which steadily reduce short-circuit strength in many US regions.

4.5 The other side of AC voltage control

AC voltage control is not automatically stable. In weak grids:

• high voltage-control gains interact with the PLL and network impedance, creating oscillations in the 1 to 20 Hz range;
• multiple devices regulating the same bus — the HVDC terminal, a STATCOM, nearby wind or solar plants — can fight unless voltage droop or coordinated settings are used;
• the PLL itself becomes less stable as SCR falls.

Grid-forming VSC control, which creates an internal voltage rather than following one, is increasingly specified for very weak connections and offshore wind hubs. In every case, EMT studies with vendor control code are required to confirm stability.

5. Inter-area oscillation damping

5.1 Inter-area modes in North America

Large interconnections exhibit inter-area electromechanical modes, typically between about 0.1 and 0.8 Hz, in which groups of generators in one region swing against groups in another. The Western Interconnection has well-known north–south modes around 0.2 to 0.4 Hz; the Eastern Interconnection has lower-frequency modes across its vast footprint. These modes are often lightly damped and are monitored with synchrophasor-based oscillation detection.

5.2 How HVDC can damp oscillations

An HVDC link in parallel with an AC corridor, or connecting two areas, can modulate its active power in phase with the oscillating speed difference between the areas, adding damping torque:

where G(s) includes washout, lead–lag compensation and limits. Input signals can be local (AC line power or bus frequency) or wide-area (synchrophasor frequency or angle differences).

5.3 Worked example: Keentel's two-area model

Keentel modeled two equal areas connected by an AC tie and a parallel HVDC link, with an inter-area mode at about 0.39 Hz.

The DC voltage stays at 1.0 pu in both cases — the Vdc terminal does its job of balancing power — but only the POD loop damps the AC mode. Damping ratios below about 3% to 5% are commonly considered inadequate in planning studies, so a POD loop can be an important system benefit of an HVDC project.



5.4 POD design considerations

5.5 Worked example: estimating POD gain from modal residues

Modal analysis of a linearized system gives each mode as an eigenvalue λ = σ + jω. A POD controller with gain K shifts the target mode approximately by:

where R is the residue of the transfer function from the HVDC power modulation input to the selected feedback signal, after phase compensation. Suppose the inter-area mode is λ = −0.047 + j2.45 (0.39 Hz, 1.9% damping) and the compensated residue is R = −0.02 + j0.01 per unit gain:

A gain giving roughly 10% damping is often a practical target, leaving margin for operating-condition changes without demanding excessive modulation. The linear estimate must be confirmed with time-domain simulation, including limits, and checked for adverse effects on other modes.

6. AC fault ride-through



6.1 What happens during a receiving-end fault

When the AC voltage at the receiving (inverter) terminal falls, its active power capability falls with it:

If the fault drops voltage to 0.2 pu and the converter prioritizes reactive current, very little active power can be exported. The sending terminal, unaware, keeps injecting power into the DC system, and DC voltage rises at a rate set by the imbalance and the DC stored energy.



6.2 Worked example: DC overvoltage in Keentel's model

The sending-end response matters: with DC-voltage droop at the rectifier, the sending AC system sees a sudden loss of export — effectively a generation surplus — which can itself be a disturbance, especially for offshore wind or weak sending systems. A DC chopper isolates the sending system from the receiving-end fault, which is why choppers are standard on offshore wind HVDC links.



6.3 Worked example: sizing a DC chopper

For a 1,000 MW link at ±320 kV (640 kV pole to pole), with a design fault duration of 150 ms at full power:

Actual designs include margins for fault duration, consecutive faults (for example, reclosing onto a fault) and the wind farm's own fault ride-through behavior.

6.4 Current limiting and reactive current priority

During AC faults, the converter's limited current must be shared between active and reactive components. For transmission-connected inverter-based resources, IEEE 2800-2022 specifies reactive current injection during low-voltage ride-through, and its scope includes VSC-HVDC systems that connect inverter-based resources. Transmission providers and ISO/RTOs specify ride-through requirements for HVDC terminals in interconnection agreements. Practical design choices include:

• reactive current priority during faults to support AC voltage;
• fast active power ramp-up after fault clearing, limited to avoid stressing a weak AC system;
• PLL design and freezing logic that maintains synchronization through phase jumps;
• negative-sequence current control for unbalanced faults.

6.5 The role of the outer loop during faults

Outer loops are slow compared with the fault event. During the fault itself, the inner current loop, current limiter, PLL and DC protection dominate behavior. Outer-loop choice matters mainly for:

• what the converter tries to do as voltage collapses (keep power, or support voltage);
• how quickly and smoothly it recovers after fault clearing;
• which terminal is responsible for DC voltage when one AC side is faulted.

6.7 VSC versus LCC fault behavior

Line-commutated HVDC relies on the AC voltage to commutate thyristor valves. AC faults and voltage dips at the inverter can cause commutation failure, momentarily collapsing DC power transfer at both ends. VSC-HVDC does not commutate against the AC voltage, so it does not suffer commutation failure and can keep supporting AC voltage during faults. When VSC and LCC links terminate near each other, the VSC terminal's reactive support can reduce the LCC link's risk of commutation failure — an interaction worth studying explicitly.

6.8 Worked example: power recovery after fault clearing in a weak system

After a fault clears, active power should ramp back quickly to limit the disturbance to both AC systems, but a step back to full power at a weak bus can depress voltage. For a terminal at SCR 1.8 transferring 0.8 pu, Keentel's model shows that recovering active power over about 0.5 s while AC voltage control supplies reactive power keeps the bus within about 1.0 to 1.07 pu. A faster ramp would reduce the energy lost during recovery but would require more reactive support and could excite control interactions. Transmission providers increasingly specify recovery times for inverter-based resources (for example, IEEE 2800 sets active power recovery expectations), and HVDC terminals should be designed to meet or exceed them.

6.9 Unbalanced faults and negative-sequence control

Most AC faults are unbalanced. During a single-line-to-ground or line-to-line fault, the converter AC bus contains negative-sequence voltage, which produces a second-harmonic (120 Hz) ripple in active power and DC voltage unless the converter controls negative-sequence current. Common strategies:

The choice affects AC line protection near the terminal — directional and distance elements rely on negative-sequence quantities — and must be coordinated with the transmission owner's protection engineers.

7. Frequency support for low-inertia systems

7.1 The island problem

An island with little synchronous generation — an isolated grid, a renewable-rich region connected by HVDC, or an offshore network — has low inertia, so a generation loss produces a high rate of change of frequency:

For a 500 MW island with H = 2 s losing 100 MW:

7.2 How HVDC provides support

7.3 Worked example: DC stored energy is not the answer

With Hdc = 0.04 s on a 300 MW terminal, DC stored energy is about 12 MJ. The island's rotating kinetic energy is 2 s × 500 MW = 1,000 MJ. DC stored energy is about 1.2% of the island's inertia — and using it means letting DC voltage fall, which the DC controls must prevent. Frequency support must come from power transferred from the main grid.



7.4 Worked example: Keentel's island model

The droop response is limited by converter headroom: a terminal already near its rating cannot help. Reserving headroom for frequency support is a planning and commercial decision. In the US, NERC BAL-003 frequency response obligations, regional under-frequency load shedding programs and ride-through requirements (NERC PRC-024 and PRC-029) define the frequency envelope that islands and their HVDC links must respect.

7.5 When there is no synchronous generation

If the island has no synchronous generators — an offshore wind collector network, for example — the HVDC terminal must be grid-forming: it sets the AC voltage and frequency for the island, and the wind turbines follow it. In that case, "frequency support" is inherent in the control: the terminal's AC frequency is a control output, and power balance is maintained through the DC link and the onshore terminal's DC voltage control.

7.6 Worked example: headroom needed for a nadir target

For a first-order estimate, the frequency nadir in a low-inertia system depends on how much power can be delivered before frequency falls to the limit. Ignoring local governors, if the HVDC delivers ΔPH within time tr after a generation loss ΔPloss, the frequency fall during the response time is approximately:

With ΔPloss = 100 MW, H = 2 s, S = 500 MW and a linear HVDC ramp to 100 MW in 0.3 s, the deficit integral is 100 × 0.3 ÷ 2 = 15 MW·s, giving Δf ≈ 60 × 15 ÷ 2,000 ≈ 0.45 Hz — close to the 0.47 Hz nadir deviation in Keentel's simulation. The HVDC link must therefore hold at least 100 MW of headroom and respond within a few hundred milliseconds to keep this island above a 59.3 Hz load-shedding threshold.

8. Choosing outer loops: a systematic approach



8.1 Decision table

8.2 Coordination questions

1. Which terminal controls DC voltage in normal operation, and what happens if that terminal's AC side is faulted?
2. How are current limits shared between active and reactive current during faults?
3. What are the AC voltage control gains and droops relative to nearby devices?
4. What supplementary controls are required by the transmission provider (POD, frequency response, runback)?
5. How do controls transition between modes (for example, after a terminal blocks)?
6. What protection settings (DC overvoltage, AC overvoltage, frequency) must coordinate with the controls?

8A. Looking ahead: multi-terminal grids and grid-forming HVDC



8A.1 Multi-terminal DC grids

Several US offshore wind and interregional proposals contemplate multi-terminal or meshed DC systems that connect multiple offshore platforms or several onshore points. In these systems, outer-loop design becomes a coordination problem:

Interoperability is an emerging challenge in North America: if each converter's controls are proprietary, integrating new terminals into an existing DC grid requires extensive multi-vendor studies.

8A.2 Grid-forming HVDC terminals

Grid-forming control is already standard for offshore wind HVDC terminals that energize and hold the offshore AC network. Onshore, grid-forming VSC terminals can provide voltage-source behavior to weak AC systems, inherent fast frequency and voltage response, and black start capability. Their challenges are the same as other grid-forming resources: current limiting during faults, coordination with protection, and interaction with nearby converters and synchronous machines. NERC has published guidance on grid-forming functional specifications, and several ISO/RTOs are evaluating grid-forming requirements for new inverter-based resources and HVDC links.

8A.3 Implications for outer-loop selection

As grid-forming control spreads, the classic outer-loop menu — P, Q, Vdc, Vac, f — evolves into a choice of synchronization philosophy (grid-following or grid-forming) plus the active and reactive objectives layered on top. The principles in this guide still apply: one or more terminals must balance DC power, damping needs a damping signal, faults are DC energy events, and frequency support depends on headroom and the strength of the remote system.

9. Modeling and study requirements

9.1 RMS versus EMT

9.2 US modeling requirements

NERC Reliability Standard MOD-032 requires data for planning models, and NERC reliability guidelines recommend EMT models for inverter-based resources and HVDC in weak areas. FERC Order No. 2023 strengthened model requirements in interconnection procedures. ISO/RTOs typically require vendor-specific EMT models (often in PSCAD) and positive-sequence models for HVDC projects, with model quality checks and benchmarking.

9.3 Model validation

Commissioning tests — step responses of each outer loop, POD tests, frequency response tests and staged fault tests where possible — should be compared with model predictions. NERC MOD-026 and MOD-027 concepts of model verification apply to controls of HVDC systems as part of the planning model data.

9.4 A practical HVDC control study workflow

1. Define the system: AC connection points, voltage classes, short-circuit levels under normal and outage conditions, nearby inverter-based resources and synchronous machines.
2. Select control philosophy: active-axis and reactive-axis outer loops at each terminal, supplementary controls and mode transitions.
3. Screen with RMS models: power flow, voltage control, frequency response, inter-area modes and POD design.
4. Detailed EMT studies: weak-grid stability, AC fault ride-through, DC faults, chopper sizing, protection coordination and control interactions with nearby converters.
5. Frequency scans: impedance-based screening for resonances and control interactions across frequency.
6. Specify: performance requirements in the converter specification — not just hardware ratings, but control behavior, response times and model deliverables.
7. Review vendor design: independent verification of vendor studies and models.
8. Commission and validate: test each outer loop and compare with models.
9. Maintain models: update after firmware or setting changes and after network changes that alter SCR.

10. Protection and control interaction

Line protection near VSC terminals deserves attention: the converter's controlled, limited fault current — and its negative-sequence behavior — can affect directional and distance elements. Protection studies with EMT models are increasingly expected.

11. Commissioning and operation



11.1 Typical outer-loop tests

11.2 Operational considerations

• Mode selection should be documented in operating procedures, with clear rules for when supplementary controls are enabled.
• POD and frequency droop require headroom, which may conflict with commercial scheduling.
• Outages in the AC network change SCR and may require different control settings; some projects implement adaptive tuning.
• Firmware changes require model updates and revalidation.

11.3 Specification checklist for outer-loop performance

Converter specifications often describe hardware in great detail and controls only briefly. Keentel recommends that specifications state, at minimum:

Clear performance requirements make vendor proposals comparable and give the owner a contractual basis to correct deficiencies.

12. Common misconceptions corrected

13. Key takeaways

1. One terminal must control DC voltage, or several must share it by droop.
2. Weak grids need AC voltage control or grid-forming, tuned with EMT models.
3. Damping needs a damping controller.
4. AC faults at the receiving end are DC overvoltage events; plan choppers and coordinated controls.
5. Frequency support is power from the other system, limited by headroom.
6. Match the study tool to the phenomenon: RMS for slow dynamics, EMT for fast ones.

14. How Keentel Engineering helps

Keentel Engineering supports transmission developers, utilities, offshore wind developers and ISO/RTO stakeholders with HVDC planning, control and interconnection studies, from 4 kV to 765 kV.

Why clients choose Keentel Engineering



• Control-aware power system engineering: we understand converter controls as well as the networks they connect.
• EMT and RMS depth: the right tool for each phenomenon.
• Practical US experience: ISO/RTO requirements, NERC standards and US voltage classes.
• Vendor-neutral perspective: independent review of vendor designs and models.
• Licensed professional engineering: a Florida-registered firm with offices in Tampa, Austin, Sacramento and Baltimore, supporting projects nationwide.

Planning a VSC-HVDC link, offshore wind export system or weak-grid interconnection? Talk to Keentel Engineering at (813) 389-7871, contact@keentelengineering.com, or book a 15-minute scoping call at calendly.com/keentel-engineering/15min.

Part C — Case Studies

The following are illustrative scenarios based on conditions typical of US HVDC projects. They show how Keentel Engineering approaches outer-loop control decisions; site details are generalized and figures are rounded.

Case Study 1 — Interregional VSC link into a weak 345 kV bus

Situation. A developer plans a 1,000 MW VSC-HVDC link delivering renewable energy into a 345 kV substation whose short-circuit capacity falls to about 1,800 MVA under credible outages — an SCR of 1.8. The initial specification calls for constant reactive power at the receiving end.

Analysis. Keentel's studies show that with fixed reactive power the receiving bus would sit near 0.85 pu at full transfer under the outage condition, outside the transmission owner's voltage criteria, and that remote faults could drive the operating point beyond the converter's capability. EMT studies with the vendor's model show that AC voltage control holds voltage but oscillates at about 8 Hz with the initial gains.

Engineered solution. Keentel recommends AC voltage control with a reactive droop coordinated with a nearby STATCOM, retuned voltage-control and PLL gains, and DC voltage control at the stronger sending terminal. Grid-forming capability is specified as an option for future network weakening.

Outcome. The revised control philosophy meets voltage criteria under all studied outages, and EMT studies show well-damped responses. The transmission provider accepts the study and the control requirements are written into the interconnection agreement.

Case Study 2 — Offshore wind export: receiving-end faults and DC overvoltage

Situation. An offshore wind project plans an 1,100 MW VSC-HVDC export system to a 345 kV onshore substation. Early studies assume the onshore terminal controls DC voltage and the offshore terminal is grid-forming.

Analysis. Keentel's EMT studies show that close-in onshore AC faults collapse the onshore terminal's export capability while the wind farm continues producing, driving DC voltage beyond protection limits within about 20 ms. Reducing wind farm output through the offshore grid frequency is too slow to prevent the overvoltage.

Engineered solution. Keentel specifies a DC chopper sized for full power for the specified fault duration, with energy capability for consecutive faults, and coordinates chopper thresholds with DC overvoltage protection and the wind turbines' frequency response for longer events.

Outcome. The system rides through all specified onshore faults without blocking, and the wind farm sees no disturbance from onshore AC faults.

Case Study 3 — HVDC link providing frequency support to a low-inertia region

Situation. A utility serving a region with high renewable penetration and limited synchronous generation relies on a VSC-HVDC link for imports. Studies of the largest credible generation loss show frequency nadirs below the first under-frequency load shedding stage when the link runs at fixed power.

Analysis. Keentel models the region's frequency response with and without P–f droop on the HVDC link. Droop sharply improves the nadir, but only when the link operates with at least 100 MW of headroom. At high import schedules, headroom disappears.

Engineered solution. Keentel recommends a 2% frequency droop with a minimum headroom requirement during low-inertia hours, combined with fast frequency response from battery storage when the link is near full import. Operating procedures define when the headroom requirement applies.

Outcome. Studies show frequency nadirs above the first load-shedding stage for all credible contingencies, and the utility gains a coordinated frequency response strategy using existing assets.

Part D — Technical FAQ: VSC-HVDC Outer-Loop Control

References and Further Reading

Links were current at publication (October 2026).



• IEEE Std 2800-2022, Standard for Interconnection and Interoperability of Inverter-Based Resources Interconnecting with Associated Transmission Electric Power Systems. IEEE.
• IEEE Std 1204, Guide for Planning DC Links Terminating at AC Locations Having Low Short-Circuit Capacities. IEEE.
• IEEE/CIGRE, Definition and Classification of Power System Stability — Revisited and Extended (2021). IEEE Transactions on Power Systems.
• CIGRE Technical Brochure 269, VSC Transmission; CIGRE Technical Brochure 604, Guide for the Development of Models for HVDC Converters in an HVDC Grid. CIGRE.
• NERC Reliability Standards TPL-001, MOD-032, MOD-026/027, PRC-024, PRC-029 and BAL-003; NERC reliability guidelines on EMT modeling and low short-circuit strength systems. North American Electric Reliability Corporation.
• FERC Order No. 2023, Improvements to Generator Interconnection Procedures and Agreements. Federal Energy Regulatory Commission.
• P. Kundur, Power System Stability and Control. McGraw-Hill.
• D. Jovcic and K. Ahmed, High Voltage Direct Current Transmission: Converters, Systems and DC Grids. Wiley.
• L. Zhang, L. Harnefors and H.-P. Nee, "Power-Synchronization Control of Grid-Connected Voltage-Source Converters," IEEE Transactions on Power Systems, 2010.
  

Disclaimer

This article is general technical information for educational purposes. It is not engineering advice for any specific project and does not create a professional relationship. Keentel Engineering's simulation results come from simplified reduced-order models with illustrative parameters; actual HVDC behavior depends on vendor-specific converter controls, protection, network conditions and operating practices and must be verified with validated vendor models, including EMT studies. Designs must follow current IEEE standards, NERC requirements, transmission provider and ISO/RTO criteria and project-specific engineering analysis, sealed by a licensed professional engineer where required.



Case studies are illustrative scenarios based on typical project conditions, not records of specific client projects. Figures are rounded and indicative.

IEEE, NERC, FERC, CIGRE, PSCAD™ and other names are the property of their respective owners. Keentel Engineering is not affiliated with or endorsed by these organizations or any equipment vendor.