Skip to content
WattCurious

How Offshore Wind Farms Send Electricity to Shore (HVAC vs HVDC)

Updated 21 September 20267 min readWind & Hydropower

Getting power from an offshore wind farm to the grid is a cable problem, not a turbine problem. A submarine cable is a long capacitor as well as a conductor: energise it with alternating current and it draws a charging current whether or not anything is connected at the far end. That current grows with length, it competes with the power you wanted to send, and at some distance it is the reason the industry stops using alternating current altogether.

Key takeaways

  • A submarine AC cable draws charging current along its whole length, so part of its rating is consumed before any wind power is carried.
  • That effect scales with length, which is why AC export becomes impractical beyond a certain distance rather than merely inefficient.
  • HVDC has no charging current in steady operation, so a DC cable's useful capacity does not fall away with distance.
  • The price of HVDC is a converter station at each end — large, complex equipment that AC schemes simply do not need.
  • The cable, not the electronics, is the usual point of failure, and a repair offshore is measured in weeks of weather windows.
On this page
  1. Everything happens inside the cable
  2. The current the cable takes before you send anything
  3. What HVDC changes
  4. Which one, and why
  5. The part that actually breaks

Everything happens inside the cable

An overhead line on land is mostly a conductor with air around it. A submarine cable is a conductor wrapped in insulation, wrapped in a metallic sheath that is bonded to the sea. That geometry — two conductive surfaces separated by a thin dielectric — is the definition of a capacitor, distributed along every metre of the route.

The electrical path from an offshore turbine to the gridA section from sea to shore. Turbines stand in rows offshore, connected in strings by medium-voltage array cables lying on the seabed. The strings run to an offshore substation platform where transformers raise the voltage. A single high-voltage export cable leaves the platform, is buried under the seabed, comes ashore at a landfall point, continues underground to an onshore substation, and connects to the transmission grid.seaseabedlandarray cables — strings of turbinesmedium voltage, a few tens of kilovoltsoffshoresubstationstep up, switch, protectone export cableburied; the whole farm depends on itonshoregridlandfallThe platform's job is to make a whole farm look like one generator with one connection point.Everything that makes offshore transmission difficult happens in the single cable that leaves it.
Strings of turbines, one platform that makes them look like a single generator, and one export cable that everything depends on.

Inside the farm the picture is conventional enough. Turbines are wired into strings by medium-voltage array cables — typically a few tens of kilovolts — because running every turbine to shore separately would be absurd. The strings gather at an offshore substation, a platform whose transformers raise the voltage for the long run and whose switchgear lets a faulted string be isolated without stopping the farm.

Everything interesting happens after that platform, in the single export cable that carries the whole farm's output. And the first thing that cable does is draw current for itself.

The current the cable takes before you send anything

Energise an AC cable and the voltage across its insulation alternates, which means the distributed capacitance is charged and discharged every cycle. That charging current flows whether or not a single turbine is turning.

Why an AC sea cable draws current for itselfThree parts. A cross-section of a submarine cable showing the conductor at the centre, insulation around it and an earthed metallic sheath outside, labelled as a capacitor formed along the whole length. A length view in which charging current flows into that distributed capacitance everywhere along the route, so the current at the sending end exceeds the current delivering power at the receiving end. A vector diagram showing the power-carrying current and the charging current at right angles to one another, with the cable's thermal rating drawn as a circle that both together must fit inside.The cable in cross-sectionearthed metallic sheathinsulationconductorTwo conductive surfaces, a thindielectric between them —a capacitor, along every metre.Along its lengthearthed sheath, bonded to the seacharging current leaks into the capacitance everywhere — with or without a farm at the far endcurrent inlessarrivesAnd they share one conductorpowerchargingtotalThe two are out of phase, so they do not simply add — butthey share the same rating, and the total is what heats the cable.Charging current rises with length, with voltage and with frequency.
The conductor and the earthed sheath form a capacitor along the whole route. Charging it consumes part of the cable's current rating before any wind power is carried.

The consequences follow directly. Charging current is proportional to length, so a longer cable draws more of it. It is also proportional to voltage and frequency, which removes the obvious escape route: raising the voltage to reduce the working current — the standard move for long distances on land — increases the charging current at the same time.

The two currents are out of phase with one another, so they do not simply add, but they do share the same conductor and the same thermal rating. Past a certain length, the cable is fully occupied carrying its own charging current and has nothing left for the farm.

Useful transmitted power against cable lengthA chart with route length on the horizontal axis and useful transmitted power on the vertical axis. An uncompensated alternating-current curve starts at full capacity and falls away with length, dropping steeply to zero once charging current has consumed the cable's rating. A second alternating-current curve, with reactive compensation added, has the same shape but reaches considerably further before collapsing. A direct-current line stays nearly flat across the whole range, declining only gently through resistive losses.Useful power deliveredRoute length →cable ratingdirect currentonly resistance to lose toalternating currentuncompensatedwith reactive compensationnothing leftnothing left, laterCompensation moves the wall further out. It does not remove it — and every reactor is more offshore structure.Illustrative shapes, not a design. Where each curve falls depends on the cable, the voltage and the compensation.
Illustrative shapes. AC capacity falls away with distance; compensation pushes the wall further out but does not remove it. DC declines only through resistance.

Reactive compensation is the mitigation: shunt reactors at one or both ends, and sometimes on a platform partway along the route, that absorb the charging current locally so it does not have to travel the full length. It works, and it is why most AC export schemes reach further than the raw physics would suggest. It also means more offshore structure to build, inspect and maintain.

What HVDC changes

Direct current does not alternate, so in steady operation the cable's capacitance is charged once and then simply sits there. No repeated charging, no charging current, no length-dependent penalty.

HVAC and HVDC export schemes comparedTwo single-line diagrams. The alternating-current scheme runs from the wind farm through an offshore transformer, through shunt reactors for compensation, along an alternating-current export cable, through more compensation onshore, into the onshore substation and the grid. The direct-current scheme runs from the wind farm through an offshore transformer into an offshore converter station that rectifies to direct current, along a direct-current cable pair with no charging current, into an onshore converter station that inverts back to alternating current, and then to the grid. Notes list what each scheme has offshore and what each pays for.offshoreonshoreHVACwind farmtransformerstep upshuntreactorshuntreactorsubstationgridAC export cablecharging current along the whole lengthHVDCwind farmtransformerstep upconverterstationAC → DCconverterstationDC → ACsubstationgridDC cable paircharged once, then nothing — no length penaltyHVAC keeps the offshore platform simple:transformers and switchgear, nothing to convert.It pays for that with a cable whose useful capacityfalls away as the route gets longer.HVDC removes the cable problem entirely:and puts power electronics on a platform at sea,for decades, with limited access — plus conversionlosses paid on every kilowatt-hour, at any distance.
HVDC removes the cable problem and adds two converter stations to do it. One of those stations has to live offshore, on a platform, for its whole service life.

Modern schemes use voltage source converters, which can set the power they transfer independently of what the grid is doing and can help support grid voltage rather than merely leaning on it. That controllability matters beyond the cable: a wind farm connected through a converter is decoupled from the grid's frequency, which changes how it behaves during disturbances.

The price is the converters themselves. Two stations, one of which is an offshore platform carrying power electronics that must run for decades with limited access. They introduce conversion losses at both ends, and those losses are paid on every kilowatt-hour regardless of distance.

Which one, and why

So the decision is a race between a fixed penalty and a growing one. HVDC pays its converter penalty immediately; HVAC pays a penalty that grows with every kilometre of route. Where they cross is commonly cited at roughly 50-100 km (30-60 mi), depending on cable design, voltage and project specifics — a range to reason with, not a threshold to design to.

What pushes a scheme towards each option
FactorPoints towards HVACPoints towards HVDC
Distance to shoreShort routes, where charging current stays modestLong routes, where AC capacity has fallen away
Farm capacitySmaller farms a single AC circuit can carryLarge farms needing full capacity at distance
Equipment offshoreTransformers and switchgear onlyAccepts a converter platform and its maintenance
LossesNo conversion losses to payLower cable losses, paid for with converter losses
Grid supportGrid must tolerate a direct AC connectionConverter can control power flow and support voltage
Connecting two gridsNot possible if they are not synchronisedStraightforward — DC does not care about frequency

Engineering relationships, not a rule. Break-even is commonly cited at roughly 50-100 km (30-60 mi), depending on cable design, voltage and project specifics — a range rather than a hard threshold, and one that also moves with the farm's capacity and what the receiving grid requires.

That last row is easy to overlook. Because a DC link imposes no frequency relationship between its ends, it can join grids that are not synchronised with each other — which is why the same technology appears in interconnectors that have nothing to do with wind.

The part that actually breaks

The converters get the attention; the cable causes the outages. It lies on or under the seabed for its whole life, and the hazards are mechanical rather than electrical.

What threatens a buried export cable, and how a repair worksA seabed section with a buried export cable and the hazards along it: a ship's anchor dragging down into the seabed, a trawl board scraping across it, a scoured section where current has washed the sediment away and left the cable unsupported and free to vibrate, and abrasion where the cable crosses exposed rock. Below, a four-step repair sequence: locate the fault, recover both cable ends to a vessel, joint them, and rebury the repaired section.water columnseabedburied to a depth chosen against the traffic overheadanchortrawl gearscoursediment washed away — the cable spans the gapfree to vibrate, and so to fatigueabrasion at a rock crossingAnd when it does fail, the farm stops exporting until a vessel can get there1 · Locate the fault2 · Recover both ends3 · Joint them4 · Rebury the sectionelectrically, from shoreto a vessel, from the seabedin controlled conditionsand survey it againEvery step waits on a weather window, and there is rarely a second route to fall back on.Which is why burial depth, route survey and periodic inspection get the attention they do — the best repair is the one never needed.Hazards are drawn on one section for compactness; a real route meets them in different places and in different combinations.
Anchors, fishing gear, scour and abrasion — all mechanical. Burial is the defence, and a repair depends on a vessel and a weather window.

Burial is the main protection, with depth chosen against the seabed conditions and the traffic overhead. Scour is the subtler hazard: currents wash sediment away from beneath a cable, leaving an unsupported span that can vibrate and fatigue. Route surveys, burial depth verification and periodic inspection exist to catch these before they become faults.

When a fault does occur there is rarely an alternative route, so the farm stops exporting until a cable vessel can locate the break, lift both ends, joint them and rebury the section — work that waits on weather in exactly the way that turbine shutdowns in storms do. It is the same theme that runs through all marine energy engineering, and the reason wave and tidal devices are judged as much on survivability and access as on output.

That is the honest summary of offshore transmission: the physics of the cable decides the architecture, and the sea decides what it is like to own. The rotor questions — how many blades, what the Betz limit permits — are settled long before any of this begins.

Frequently asked questions

Why not just use a higher voltage for AC and be done with it?

Raising voltage reduces the current needed for a given power, which helps with resistive losses — but charging current rises with voltage as well, so it makes the capacitive problem worse rather than better. That is the trap: the usual remedy for distance on land works against you at sea.

What is reactive compensation?

Equipment — typically shunt reactors — that absorbs the cable's charging current locally instead of letting it travel the length of the cable. It is installed at one or both ends, and sometimes at a platform partway along. It extends the workable length of an AC link, at the price of more offshore structure to build and maintain.

Does HVDC lose less energy?

The cable itself does, because there is no charging current and no skin effect to contend with. The converter stations at each end introduce losses of their own that an AC scheme never pays. So HVDC is not automatically more efficient — it wins on long links, where the cable advantage has had enough distance to outgrow the fixed converter penalty.

What is inside an offshore substation?

Transformers to raise the array voltage to export voltage, switchgear to isolate faults and individual strings, protection and control systems, and for a DC scheme the converter equipment as well. It is a platform whose job is to make the whole farm look like a single connection point to the export cable.

What happens when an export cable fails?

The farm stops exporting, entirely, until it is repaired — there is usually no second route. Locating the fault, mobilising a cable vessel, recovering the cable from the seabed, jointing it and reburying it depends on weather windows, so outages are measured in weeks rather than hours. This is why burial depth and route surveys receive so much attention.

Sources

Named organisations whose published material underpins this article. Where no link is given, the source is named rather than linked.

Editorial Team

Research, drafting and review

Articles are drafted from primary engineering and physics references with AI-assisted tools, then reviewed and fact-checked line by line by a human editor before publication. We publish explanations, not recommendations: no products, no pricing, no country-specific rules, and no invented author personas.

Last reviewed 21 September 2026. How we research and review

Wind & Hydropower

What Is the Betz Limit?

No open-flow rotor can capture more than 59.3% of the wind's power. The derivation in full, what the limit does not say, and where it does not apply.

Updated 21 September 20267 min read