Safety, Fire & Maintenance

Ordinary building electrics run on alternating current from the grid, and you can isolate them at a switch. A PV array is different in two ways that make it more hazardous, and understanding both is the foundation of everything else. First, it is a source, not just a load: the modules themselves generate power, so switching off the inverter or the main supply does not de-energise the array. The DC wiring from the modules stays live whenever there is light. You cannot make a PV roof electrically dead by flipping the building's main switch - the sun is the switch, and it is on.

Second, PV arrays produce direct current at high voltage. Modules are wired in series into 'strings' whose voltages add up, so the DC conductors between the array and the inverter can sit at voltages high enough to be lethal and, critically, to sustain a dangerous arc. DC arcs are especially hazardous: unlike AC, whose current passes through zero a hundred times a second and tends to self-extinguish an arc, DC flows steadily and an arc, once struck - across a loose connector, a damaged cable, a corroded joint - can persist, burn intensely hot, and start a fire. Faulty connectors and poor installation are a well-known cause of PV fires precisely for this reason.

For an integrated envelope, this is sharpened further. In BIPV the live DC wiring is not on a rack above the roof but woven into the fabric of the building - behind facade panels, above ceilings, within the construction - which can make it harder to inspect, harder to isolate a fault, and harder for firefighters to know what is live and where. None of this makes PV unsafe to have; well-designed, well-installed, code-compliant systems operate safely all over the world. But it explains why PV safety cannot be treated like ordinary wiring: the array is a permanent, self-energising, high-voltage DC source built into the building, and that demands specific protective design - arc-fault detection, proper connectors and installation, correct isolation, clear labelling of what is live - all of which is the electrical engineer's binding work to the governing codes, never a matter of convenience or a designer's assumption.

Because the array will not switch off, fire safety for PV has two linked concerns: reducing the chance of a PV-started fire, and protecting the firefighters who must tackle any fire in a building that has one. The first is largely about quality and protection - certified modules and equipment, proper connectors, good installation, arc-fault protection - the engineer's and installer's domain. The second introduces ideas a designer should understand by name.

Rapid shutdown is the principle that, in an emergency, the dangerous DC voltage on and around the array can be quickly reduced to a safe level, so that firefighters and workers are not facing live high-voltage conductors across the whole roof. Depending on the system and the code, this can be achieved at the array boundary or right down at each module (module-level electronics), shrinking the live zone. The exact requirement - whether rapid shutdown is mandated, to what voltage, over what area, how fast - is set by the governing electrical and fire codes, not chosen for convenience, and its design belongs to the engineer. Firefighter access is the architectural counterpart: a roof covered edge-to-edge in modules is a roof a fire crew cannot cross, cut ventilation through, or work on. Codes commonly require clear pathways and setbacks around and between arrays so firefighters can reach, move and vent - and honouring that is a genuine design decision the architect makes early, because it shapes the array layout and therefore the envelope.

Two more realities matter. Labelling and signage - clear marking of live DC, disconnects and shutdown points so responders know what they face - is required and life-saving. And the envelope material itself: on facades especially, and on tall buildings above all, the fire performance of the whole assembly (the PV element and everything behind it) is a serious matter governed by fire codes and the fire authority, with the tragic history of combustible cladding as the backdrop. For a designer, the takeaways are concrete: design the array layout to preserve firefighter access from the outset, coordinate space for shutdown and clearly-labelled isolation, and treat the fire performance of an integrated PV facade as a binding question for fire engineers and the authority - never as a purely aesthetic or cost choice. The specifics are theirs; the early design accommodation is yours.

The same properties that make a PV envelope a fire concern make it a hazard to work on, and it will be worked on - during installation, inspection, cleaning, repair and eventual replacement - across its whole life. Two hazards stack. There is the electrical hazard: the array is live in daylight and carries high-voltage DC that cannot be switched off by a normal isolator, so anyone working on or near it risks shock and arc unless the system is properly isolated (as far as it can be) and safe procedures are followed. And there is the height and access hazard: the work happens on roofs and facades, so falls, working at height, and access difficulty are ever-present - a facade-integrated module several storeys up is a serious access problem quite apart from its wiring.

This is why work on PV systems is emphatically for qualified, trained people using the right procedures and equipment - safe isolation as far as possible, appropriate personal protective equipment, fall protection, and an understanding that the array may still be live. It is not DIY, and it is not general maintenance-staff work; it is specialist work governed by electrical-safety and working-at-height rules. A designer does not write those procedures, but two design consequences follow directly and are firmly the designer's responsibility.

First, design for safe access from day one. An integrated module that cannot be reached safely to inspect, clean, repair or replace is a liability that will haunt the building - so access provisions, anchor points, walkways, and reachable detailing must be planned into the envelope, not bolted on as an afterthought when someone first needs to get to a failed panel four floors up. Second, respect the maintenance reality in the design of the integration itself: how a BIPV element is fixed, sealed and connected determines whether a single failed unit can be replaced without dismantling half a facade or breaching the waterproofing. Getting these right is architecture. But the safety of the working practices, the isolation procedures, and the electrical and fall-protection design are binding matters for the electrical engineer, the installer and the relevant safety codes - understood and accommodated by the designer, decided by the specialists.

A PV envelope is a decades-long relationship, not a fit-and-forget finish, and designing as if it were the latter is a quiet but serious failure. A well-made system is genuinely low-maintenance, but 'low' is not 'none', and over twenty-five or thirty years several things reliably need attention. Cleaning and soiling: dust, dirt, bird droppings, pollen and, in Indian conditions especially, heavy dust can build on modules and cut output noticeably; periodic cleaning restores it, and how easily that can be done safely is set by the design. Monitoring: modern systems report their output, so a drop from a failed module, a tripped string or a dying inverter can be caught rather than silently costing generation for months - monitoring turns invisible faults into visible ones, and is a real part of owning a generating building.

Component replacement: the modules may last decades, but the inverter is the shorter-lived heart of the system and typically needs replacing at least once (often around the ten-to-fifteen-year mark) within the building's life - a cost and an access requirement that belong in any honest plan. Batteries, if fitted, are replaced too (previous lesson). Degradation: modules slowly lose output over the years - a gradual, expected decline, not a fault - so a system's generation late in life is somewhat less than when new. And for BIPV specifically, the module is also building fabric: its weatherproofing, sealing and fixing must keep performing as building envelope for the whole life, so maintenance is not only electrical but also about the integrity of the skin.

For the designer, the maintenance reality translates into decisions made at design time, illustrated in the timeline figure: plan safe, practical access for cleaning and inspection; ensure the inverter and other serviceable parts are reachable and replaceable without major disruption; detail integrated modules so a single failure can be addressed without wrecking the facade or its waterproofing; and set honest client expectations about upkeep, monitoring, inverter replacement and gradual degradation over the system's life. All the specific figures here - cleaning intervals, inverter lifespan, degradation rates - are illustrative and depend entirely on the products, the site and the climate; the binding maintenance requirements and safety procedures come from the manufacturers' data, the electrical engineer and the governing codes. What the designer owns is designing a generating envelope that can actually be looked after, safely, for the decades it will live.