A DC Choice Inside an AC Grid
A converter station designed to move several gigawatts between countries can fail to switch if the AC voltage on the receiving end dips too low. That failure mode sits at the heart of the HVDC LCC vs. VSC converters comparison: it separates line-commutated converters, which depend on the AC network they connect to, from voltage-source converters, which can establish their own. It also explains why a long-distance hydro exporter chooses a different architecture than an offshore wind developer. HVDC converter stations therefore cost differently, provide different grid services, and fail in different ways. Understanding where each type wins is less about semiconductor preference than about network conditions, required services, and what the project is actually for.
Alternating current dominates almost every synchronous grid because transformers make voltage conversion inexpensive and familiar. HVDC typically earns its place where AC’s limits dominate: very long overhead or cable distances, submarine crossings, and interconnections between grids that operate at different frequencies or cannot be synchronised. Those conditions are the classic drivers, though controllability, the grid services a link is expected to provide and the underlying project economics can also favour a converter-based solution.
Over long AC cables, capacitance charges and discharges with each cycle, generating reactive current that consumes part of the cable’s thermal capacity. Beyond a certain distance, that charging current can make AC cable transmission electrically impractical, which is why submarine links often become the first economical use of HVDC within the high-voltage grid.
Early HVDC links used mercury arc valves, a predecessor to the thyristors that now define LCC designs. Those first projects were built to move hydroelectric output across water where AC cable capacitance posed a practical problem. That history matters because it shows HVDC was never a general replacement for AC. It is a targeted tool for distances and boundaries where AC’s electrical characteristics or economics become unattractive.
The LCC/VSC distinction emerged as power electronics advanced. LCC designs use line-commutated thyristors, while VSC designs use insulated-gate bipolar transistors, commonly called IGBTs, which can be switched on and off independently. That one difference in switching capability runs through every subsequent trade-off: converter cost, loss profile, station footprint, and how the link behaves when the surrounding AC network is weak or has no generation of its own. It also determines what each topology asks of the grid at either end.
Line-Commutated Converters: The Workhorse of Bulk Transmission
Line-commutated converters use thyristors. A gate signal can turn a thyristor on, but it cannot turn it off. Instead, the AC voltage at the connected bus reverse-biases the device and forces it to stop conducting. That is the meaning of line-commutated: the converter borrows its turn-off signal from the external AC system rather than controlling it internally.
The practical consequence is that an LCC terminal needs a reasonably strong AC network at both ends. If the connected AC voltage drops too low, commutation can fail, causing the converter to misfire or block. During a commutation failure, DC current may briefly transfer to a bypass path, and the link may need to restart or ride through the disturbance. This constraint directly shapes where LCC can be deployed, and it explains why some existing links require careful short-circuit strength assessment as the surrounding grid changes.
LCC cannot control its reactive power independently. Every LCC terminal consumes reactive power from the AC grid, typically a large share of the active power being transferred. Converter stations therefore include AC filter banks, shunt capacitors, or additional static var compensators to supply that reactive demand and to absorb the harmonic currents produced by line commutation.
Where those requirements are manageable, LCC offers clear advantages. It is the established technology for ultra-high-voltage direct current links, particularly bulk hydro and remote generation corridors where generation sits far from load. The power ratings and distances involved are project- and system-specific rather than fixed, but links of this kind are typically built at multi-gigawatt ratings over distances measured in hundreds of kilometres and more.
LCC converter losses tend to be lower than VSC losses at equivalent voltage and power, and its converter station costs are typically lower per megawatt because thyristor technology is older, cheaper, and available at very high current and voltage ratings. For very long point-to-point links between strong AC networks, LCC remains the default. China’s long-distance UHVDC corridors are a prominent example of this application.
At the same time, LCC’s dependence on a strong AC grid makes it a poor choice for isolated renewable plants, offshore wind farms, or weak island systems. It also cannot black start a network: without an AC voltage to commutate against, the converter cannot operate. In many modern grid reinforcement applications, those limitations point toward VSC.
Voltage-Source Converters: Flexibility at a Higher Price
Voltage-source converters use insulated-gate bipolar transistors, which are controllable switches that can be turned on and off at high frequency. This self-commutation removes the dependence on an external AC voltage. A VSC can operate into a weak grid or a dead network, and it can control active and reactive power independently across all four quadrants of its capability curve.
This capability changes how the converter functions within the grid. Instead of consuming reactive power, a VSC terminal can be programmed to supply or absorb reactive power to support voltage at its point of connection. The same converter can provide synthetic inertia or fast frequency response, services that become more valuable as synchronous generators retire. In the absence of a live AC grid, a VSC can use its own switching to establish an AC voltage, making black start possible from the converter itself.
The trade-off is cost and complexity. IGBT-based valve designs are more expensive per megawatt than thyristor valves, and the switching process introduces higher losses, although modern modular multilevel converter topologies have reduced those losses considerably. VSC stations also tend to have a smaller physical footprint for a given rating because they require fewer passive AC filters, but the power electronics themselves represent a higher capital cost.
Offshore wind is one of the clearest commercially mature applications for VSC. Submarine HVDC links connecting distant offshore wind farms to onshore grids routinely use VSC because the offshore AC network is weak, platform space is limited, and the converter must manage voltage and frequency without relying on a strong onshore grid. Multi-terminal HVDC systems and interconnections between asynchronous grids also favour VSC.
Where the Comparison Actually Bites
For a planner or developer, the LCC-versus-VSC decision rarely comes down to a preference for one semiconductor. It comes down to AC network conditions, the required grid services, the distance, the power rating, and the acceptable capital and loss profile. A technology choice made decades ago can also interact with the changing generation mix. A region that built LCC links around large thermal plants may find those links more challenging as those plants close and rotating inertia declines.
The choice becomes more complicated when the AC network is not simply strong or weak. A converter terminal may sit in a network that is normally strong but loses short-circuit strength during outages or as synchronous machines retire. In that situation, an LCC link that once performed well can face repeated commutation failures or require additional synchronous condensers. Several existing HVDC links have been retrofitted with synchronous condensers or dynamic reactive support for exactly this reason.
Grid codes add another layer. Some jurisdictions now list specific service requirements for HVDC links, particularly frequency response, fault ride-through and voltage control. LCC can be adapted with additional equipment, but self-commutated VSC can provide several of these services directly. Where a converter is being built to comply with modern grid codes, the required services can override a purely cost-driven comparison.
- AC strength: LCC needs a strong AC grid at both ends; VSC can operate in weak networks and start against a dead bus.
- Reactive power: LCC consumes reactive power and requires filters and compensation; VSC can supply reactive power and support voltage control.
- Harmonics: LCC generates low-frequency harmonics that require large AC filters; VSC uses higher-frequency switching and generally needs less filtering.
- Losses and cost: LCC tends to have lower converter station losses and lower capex per MW for large point-to-point links; VSC tends to cost more but avoids some AC reinforcement and filter costs.
- Reliability behaviour: LCC is mature and well understood, but commutation failures can occur during AC disturbances; VSC avoids commutation failures but relies on more complex controls and power electronics.
The interaction between converter choice and grid evolution is one of the less visible consequences of long planning timelines. Converter types selected today need to function in a grid that may differ materially from the one that justified them. Planning studies often examine short-circuit strength under future outage conditions as well as present system conditions.
What the Choice Means for Utilities
Both converter families remain relevant, but in different parts of the grid. VSC costs have declined as IGBT manufacturing matured and modular multilevel converter designs standardised, making VSC viable for a broader set of interconnectors and offshore links. LCC retains an edge for very high-power, very long-distance overland corridors where the AC network at both ends is strong and the primary objective is moving bulk energy at low losses.
Hybrid and multi-terminal arrangements are beginning to blur the boundary. Some converter station designs combine LCC and VSC terminals on the same DC system, using LCC for the bulk transmission stage and VSC where weak AC conditions or black start capability are needed. Hybrid designs assign each technology to the role it handles best rather than removing the underlying differences.
For many projects, the semiconductor choice is less pressing than the surrounding infrastructure. Converter stations, DC cable manufacturing capacity, and HVDC switchgear all face supply-chain constraints. After a technology is selected, transmission infrastructure still has to be permitted, manufactured, and installed on schedules that rarely match the speed of generation connection requests. HVDC does not compress those planning timelines; in some cases, its specialist equipment lengthens them. Cost allocation adds another layer, because the benefits of a long HVDC link are rarely confined to a single utility or market participant.
References
- CIGRE Technical Brochure 492 — Voltage Source Converter (VSC) HVDC for Power Transmission — Economic Aspects and Comparison with other AC and DC Technologies. Used for converter loss, cost, and application comparisons.
- IEEE Std 1204-1997 — IEEE Guide for Planning DC Links Terminating at AC Locations Having Low Short-Circuit Capacities. Used for AC strength and commutation planning considerations.
- CIGRE Technical Brochure 139 — Guide to the Specification and Design Evaluation of AC Filters for HVDC Systems. Used for reactive compensation and harmonic filtering context.
- ENTSO-E Network Code on HVDC Connections, Commission Regulation (EU) 2016/1447. Used for grid service and frequency support expectations for HVDC converter stations.
