11.3 DC Cables and Earth-Fault Causes

Quick Answer: Handbook v1.3 Learning Outcome 4.2.5(c)–(d) requires DC cables to be chosen for the external influences of the route. Cables behind a PV array need a temperature rating of at least 70 °C. Avoid burying those cables in walls where possible. Provide identification and labelling, and keep circuit length down to minimise voltage drop. Common DC earth faults come from water ingress, mechanical damage (including animals and the environment), and UV or heat degradation. Typical UK PV DC cable is the BS EN 50618 family, often marked H1Z2Z2-K — that is IET Code of Practice / industry practice, not a number printed in the 2922-34 handbook extract. The label PV array cable – live during daylight is physically true. This OpenExamPrep section is independent study material for those points; it is not a City & Guilds publication.

External influences decide the cable, not the merchant’s leftover drum

Handbook 4.2.5(c)(i) is the selection sentence electricians already know from BS 7671: match the wiring system to the influences. On a photovoltaic DC route those influences are harsher than a plastered ring final.

Temperature. The air gap behind modules on a still, sunny roof is a hot enclosure. Laminates run well above 25 °C cell temperature; the cable sitting on felt or against a membrane sees elevated ambient. Ordinary 70 °C PVC is already at the handbook minimum behind an array and is usually the wrong product for UV as well.

Solar radiation (UV). Roof trays, eaves, and unguttered drops see years of UV. A cable that chalks, cracks, and then lets water onto the copper is an earth-fault machine.

Water. Rain, snow, ponding trays, valley gutters, and wind-driven rain at glands. DC earth faults are often wet tracking from a damaged sheath or a flooded connector back to an earthed tray, rail, or wet roof covering.

Flora and fauna. Squirrels, rats, nesting birds, and insects. Chew marks on a DC pair in a loft or at an eaves soffit are not a curiosity. They are a live-during-daylight earth fault waiting for a damp day.

Mechanical. Tile edges, wind-induced movement, foot traffic on a “finished” roof, stored loft items crushing a run, abrasion on a brick corner, impact from a ladder. PV cable is flexible, not armour unless you specified an armoured or ducted system for that influence.

Other. Farm ammonia, coastal salt, industrial dust, and heat from a flue or an unventilated south wall all belong in the same selection thought. The handbook does not print a catalogue. It requires you to notice the environment the cable will actually occupy.

At least 70 °C behind the array

Handbook 4.2.5(c)(ii) is a hard minimum: cables routed behind a PV array should have a temperature rating of at least 70 °C. That is not a recommendation to use 70 °C PVC singles because the number matches. It is a floor. Competent UK practice for exposed DC is a photovoltaic cable designed for high conductor temperature, UV, and ozone. The usual product family in the IET Code of Practice for Grid-connected Solar Photovoltaic Systems and in merchant listings is BS EN 50618, often marked H1Z2Z2-K (tinned Class 5 copper, cross-linked insulation and sheath, typically dual-insulated single cores). BS EN 50618 / H1Z2Z2-K is not a number printed in the 2922-34 handbook extract. Do not invent it as a handbook citation on the paper. Do use it as the usual way to meet the handbook’s temperature, UV, and mechanical influences on a roof. Typical industry ratings for that family sit well above 70 °C conductor temperature; always read the datasheet for the drum you pull.

A 4 mm² or 6 mm² PV single in a ventilated tray behind the modules is the normal 2922 picture. A 6 mm² 6491X or twin-and-earth stuffed into that gap because it was on the van is not “at least 70 °C and UV-ready.” It is the wrong influence match even if the copper cross-section looks generous.

Avoid burying in walls where possible

Handbook 4.2.5(c)(iii) is blunt: avoid burying in walls, where possible. A buried PV DC circuit is a generator conductor that stays live in daylight, cannot be inspected, cannot be withdrawn when a connector fails, and can be nailed, chased, or overheated in insulation without anyone seeing the damage. If a wall penetration is unavoidable, treat it as a designed crossing: a duct or containment from which the cable can be withdrawn, identified, and replaced — not a plastered-in live pair that someone will later assume is a dead lighting radial.

The same thought applies to loft insulation. Packing DC singles into quilt is burying them in a hot, inaccessible, mechanically hostile place. It is not clever concealment from ordinary persons. It is how animal damage and overheating become hidden earth faults in an occupied dwelling.

Identification, labelling, and why live during daylight is physically true

Handbook 4.2.5(c)(iv) requires identification and labelling. Polarity must be obvious for the whole life of the array: a reversed DC pair is a later testing chapter, but the cable run is where the reverse is built. String identity matters in a combiner. The later labelling outcome gives the wording PV array cable – live during daylight (or equivalent). This section teaches why that sentence is physically true, not the full label schedule.

A PV module is a semiconductor current source. Photons generate charge as soon as daylight — including bright overcast, low winter sun, and strong albedo from snow or pale render — reaches the cells. There is no off switch on the laminate. Opening a downstream DC switch-disconnector isolates the inverter. It does not stop Voc appearing on the array side. The copper in the tray, the loft, or the wall chase is still a live generator conductor whenever the modules can see light. That is why a person who opens a loft hatch at 10:00, or who chisels a wall five years later looking for a “spare” cable, can meet several hundred volts DC without anyone having “turned the solar on.” Night reduces generation; it is not a work method, and hybrid plant can still energise DC from a battery. Street lighting and strong reflections can produce voltage at odd hours. The default the label encodes is: if this is a PV array cable, treat it as live in daylight.

Do not rely on colour alone. Do not leave a DC pair looking like a brown-and-blue AC circuit. Do not put the live-during-daylight label only at the inverter and leave twenty metres of unmarked roof tray for the next trade.

Circuit length and voltage drop

Handbook 4.2.5(c)(v) asks you to consider circuit length to minimise voltage drop. Drop on DC is I × R along the pair. It wastes watt-hours as heat in the cable. It also steals voltage from the inverter’s maximum power point window: a long, thin run can leave a cold-morning string above the inverter’s maximum while a hot afternoon string sits unnecessarily low after the drop. The design answer is short DC (inverter close to the array where siting allows) and adequate cross-section, not a heroic 40 m of 2.5 mm² because the drum was open.

The AC-side chapter quotes a handbook figure of not more than 1 percent where practicable for AC voltage drop. Do not transplant that percentage onto DC as if 4.2.5 printed it. The DC instruction in this outcome is to minimise drop by length (and, in competent design, by CSA). Record the calculated drop in the design pack against the inverter’s voltage window. If the only way to keep DC short is a different inverter location, that is an equipment siting decision, not a cable you hide in a wall to “make it neat.”

Common causes of DC earth faults

Handbook 4.2.5(d) lists three causes. They are how insulation from a live DC conductor to earth (rail, tray, wet covering, metallic enclosure) fails.

Water ingress. Flooded connectors, failed glands, cracked sheath, a combiner with the wrong IP, capillary up a hanging tail with no drip loop. Water lowers insulation resistance, corrodes contacts, and provides the tracking path. Many “mystery” inverter insulation alarms are wet DC, not a failed laminate.

Mechanical damage to cables (animal, environmental). Chewed sheath in a loft or soffit. Abrasion on a tile. Wind fatigue where a loop slaps a bracket. Impact from stored boxes. Environmental includes UV-cracked sheath that then tears, and thermal expansion that pulls a gland until the seal opens. The handbook names animal and environmental together so you do not treat squirrels as outside the electrical design.

UV or heat degradation. The wrong polymer on the roof. Cable behind modules below the 70 °C thought. A black enclosure that cooks terminations. Heat plus UV makes PVC brittle; the next wind or animal finishes the earth fault.

These three cause earth faults because the copper, once exposed, finds an earthed path: a bonded rail, a wet membrane, a metallic tray, a damper loft tank. The inverter’s insulation monitor (the next chapter) may alarm. The cable choice and route are how you prevent the fault, not how you wait for a light to flash.

Influence versus cable choice

Scenario: 6 mm² singles in loft insulation versus UV-rated tray on the roof

Two quotations for the same 4 kWp south roof. Both use 6 mm² copper, so both claim “no voltage drop problem.”

Quote A drops indoor-style singles through the loft, buries them in quilt “for neatness,” and pops through a wall chase to a garage inverter. The copper size may limit ohmic drop on a short house, but the influences are wrong: heat in insulation, no UV story if any part sees a soffit gap, no inspectable route, animal chew hidden until an earth fault, and a buried live-during-daylight circuit in the wall. That is 4.2.5(c) failed even if the millivolt drop looks pretty.

Quote B uses UV-stable PV cable in a drained roof tray, drip loops at every connector, a protected drop in containment, and a short, labelled DC pair to the inverter. Behind the modules the temperature rating meets the 70 °C floor (and, in normal UK practice, the BS EN 50618 family). Voltage drop is managed by length as well as CSA. Earth-fault causes — water, mechanical/animal, UV/heat — have been designed against instead of insulated over.

The exam trap is treating 6 mm² as a complete DC design. Cross-section is only one variable. Influence, temperature behind the array, burying, identification, and length are the 4.2.5 list. The live-during-daylight label on Quote B is not decoration. It tells the next person in that loft the physics that is still true whenever the sun is up.