HVDC and Polymeric Insulation

Alternating current won the "war of the currents" because the transformer let it change voltage effortlessly, and for a century AC has carried the world's power. Yet direct current never lost its case for a few special tasks, and modern power electronics have handed it a triumphant return. HVDC excels where AC struggles. Over very long overhead lines it carries more power with lower losses and needs no reactive compensation. In cables — submarine or long underground runs — the advantage is decisive: an AC cable draws a continuous capacitive charging current that grows with length until it consumes the cable's whole rating, capping useful AC cable length at a few tens of kilometres, whereas a DC cable draws no steady charging current and can run for hundreds of kilometres. HVDC also forms an asynchronous tie between grids of different frequency or phase that cannot be joined by AC, and its converters give fast, precise control of power flow.

The price is the converter station at each end. Older line-commutated converters (LCC) use thyristors and are robust and efficient for bulk point-to-point links; modern voltage-source converters (VSC), built from IGBTs in modular multilevel arrangements, are compact, can start a dead network, control reactive power independently, and — as we will see — sidestep the brutal polarity-reversal stress of LCC. VSC has opened HVDC to offshore wind farms and to multi-terminal DC grids, making direct current a cornerstone of the renewable-era network.

Here is the conceptual heart of the chapter. Under AC, the steady-state field inside a dielectric is governed by permittivity \(\varepsilon\): the insulation behaves as a capacitor, and the field grades according to the geometry and \(\varepsilon\), which barely changes with temperature. Under DC in steady state, the displacement currents have died away and the field is instead governed by conductivity \(\sigma\): the insulation behaves as a (very poor) resistor, and the field grades so that the conduction current density \(J=\sigma E\) is continuous.

The twist is that conductivity is far more sensitive than permittivity — it rises strongly with both temperature and field, often by orders of magnitude. So in a loaded DC cable, whose conductor is hot and whose sheath is cool, the inner insulation becomes much more conductive than the outer. Continuity of current then forces the field down in the hot, conductive inner region and up in the cool outer region — the maximum stress migrates from the conductor (as in AC) out to the sheath. This field inversion with load has no AC analogue, and it means a DC cable's worst-stressed point depends on how hard it is working — a profound complication for design.

Field inversion is only half the DC story; the other half, and the deeper one, is space charge. Under sustained DC stress, charge carriers are injected from the electrodes and generated within the bulk, and because a polymer is full of traps, they do not flow straight through but accumulate inside the insulation. This trapped charge has its own field, which adds to the applied field — and where it piles up, the local field can climb far above the average, even though the terminal voltage is unchanged.

Space charge is the reason extruded polymer cables, mature and excellent for AC, were for years considered unsuitable for DC: the accumulated charge produced unpredictable field enhancements that aged the insulation and could trigger breakdown. Two advances changed this. First, the field can now be seen: the pulsed electroacoustic (PEA) method probes the charge profile through the insulation thickness, turning an invisible problem into a measurable one. Second, the charge can be tamed: nanodielectrics — XLPE filled with a few percent of nanoparticles such as magnesium oxide — introduce a fine distribution of traps that limits carrier mobility and suppresses charge accumulation, holding the field close to its designed value. These "space-charge-suppressed" materials are what finally made extruded DC cables practical, and they are an active research frontier.

The polymer at the centre of all this is cross-linked polyethylene (XLPE). Ordinary polyethylene is a fine dielectric but softens when warm; cross-linking its molecular chains (with peroxide or silane) locks them into a network that keeps its shape up to a continuous conductor temperature of \(90^{\circ}\mathrm{C}\), with very low loss and no oil to leak or maintain. Extruded XLPE has all but replaced the older oil-impregnated paper and mass-impregnated cables for AC, and — thanks to the space-charge-suppressed grades above — is now taking over DC as well.

The structure repays study. The stranded conductor is wrapped in an inner semiconducting screen that smooths the field so no strand edge concentrates stress; then the XLPE insulation; then an outer semiconducting screen, a metallic sheath for earth return and moisture sealing, and a protective jacket. The cable is made by extruding insulation and both screens together (triple extrusion) and curing them dry, because the enemies of XLPE are contaminants, voids and moisture: a speck of dust or a microvoid becomes the seed of the ageing process we turn to next.

Polymeric insulation ages chiefly by treeing — the slow growth of tiny, branching channels that resemble trees and that march, over years, across the insulation until they bridge it. Two kinds matter.

An electrical tree begins at a void, contaminant or protrusion where the field is locally intense enough for partial discharge; each discharge erodes a little more polymer, extending a fine channel, which branches and advances until it crosses the insulation and causes breakdown. A water tree is gentler and slower: in AC cables exposed to moisture, diffuse bush- or fan-shaped growths form at much lower stress, not by discharge but by the combined action of water and field; they rarely break the cable directly but degrade the polymer and often become the birthplace of an electrical tree. Unlike the air gaps of earlier chapters, a treed polymer does not heal — treeing is the irreversible clock that sets the service life of a polymeric cable, which is why manufacturing cleanliness and moisture exclusion matter so much.

Polymers have transformed the line as well as the cable. For a century, overhead lines hung on porcelain and glass insulators — strong and durable, but heavy, brittle and prone to flashover when their surfaces grow wet and dirty. Modern lines increasingly use composite (polymeric) insulators: a slim fibreglass-reinforced (FRP) core rod that carries the mechanical load, sheathed in silicone-rubber housing and weather sheds. They are a fraction of the weight, resist vandalism (they do not shatter) and are easier to install.

Their decisive advantage is hydrophobicity. Silicone rubber repels water, so rain and dew bead into separate droplets rather than spreading into the continuous wet film that carries the leakage current and triggers pollution flashover on a porcelain string. Remarkably, silicone even transfers its water-repellency to a layer of surface pollution and recovers it after a soaking, keeping the surface effectively dry under conditions that would flash over ceramic. The cost is that the polymer housing itself ages — by tracking and erosion under leakage current, and by UV and weather — so material formulation and adequate creepage distance remain essential, especially in the heavily polluted or coastal environments where the long surface path resists flashover. Under DC the pollution problem is worse still, because the unidirectional field steadily attracts charged contamination to the surface.

The converter station concentrates the hardest insulation duties of an HVDC scheme. The converter transformer, the valves, the wall bushings and the DC yard all see a mixture of AC and DC stress superimposed, and their insulation must coordinate both at once — a far less settled art than the AC coordination of Chapter 25. One stress is peculiarly severe and peculiarly instructive: polarity reversal.

In a line-commutated scheme, reversing the direction of power flow means reversing the voltage polarity. But the insulation is full of space charge built up under the old polarity, and that trapped charge does not vanish instantly. When the applied field suddenly flips, the slow space-charge field now adds to it instead of opposing it, and the total field briefly soars — a transient stress that can exceed anything the steady DC voltage would suggest. This is a direct, practical consequence of the space-charge physics of Section 26-3, and it is the reason LCC converter transformers and cables are designed and tested specifically for polarity-reversal duty. It is also a major reason VSC schemes are attractive: they change power direction by reversing the current while holding the voltage polarity fixed, so the insulation never suffers a reversal at all.

Stand back and the modern grid's materials story is clear. Polymers — XLPE in the ground and under the sea, silicone on the towers — have displaced the oil-paper and porcelain of the classical era, trading maintenance and weight for new failure modes (space charge, treeing, tracking) that the engineer must understand rather than avoid. HVDC, rebuilt on power electronics, ties continents and harvests offshore wind, but only because space-charge-suppressed nanodielectrics finally let a polymer hold a clean DC field. The research frontier runs straight ahead from here: nanodielectrics engineered trap by trap, recyclable thermoplastic cable insulation that needs no cross-linking, and the condition monitoring of polymer assets that Chapter 21 began.

This also opens the rest of Part 8. The same fast, high-voltage, polymer-insulated technology that switches a VSC underlies the pulsed-power systems of the next chapter and the pulsed-electric-field applications beyond it, where microsecond high-voltage pulses are put to work in industry, medicine and the environment — and where the smart, data-driven asset management of the final chapter keeps the whole modern apparatus healthy.

For each item, first identify what it tests — the case for DC, the conductivity field and inversion, space charge, XLPE structure, treeing, composite insulators, or polarity reversal — then apply it. Difficulty rises down the list.