Anatomy of a PV System

Follow the power in order and the system stops being a mystery. It begins at the modules - the assembled PV cells on the sunlit surface. Under light they generate direct current (DC): a steady, one-direction flow at a voltage that rises as you wire modules in series into a 'string'. This is the only place generation actually happens; everything downstream is about moving, converting and controlling that power without wasting or endangering it.

From the modules the DC travels through DC cabling - purpose-made, UV-stable, double-insulated cable - to a DC isolator, a switch that lets the DC side be shut off for safety and maintenance. This DC side is important to respect: it is live whenever the sun shines, cannot simply be switched off at the modules, and sits at a voltage that demands proper design. Next comes the inverter, the brain and workhorse of the system, which converts DC into alternating current (AC) - the form the building's wiring and the grid use - and does so while constantly hunting for the point that extracts the most power (more on that in 2.2).

The AC output passes through AC protection (breakers, isolators, earthing) and reaches the building's consumer unit (the distribution board). Here the solar AC mingles with the building's ordinary supply: it first feeds whatever the building is using right now, and only the surplus continues on. A meter records the flows - in a net-metered installation it counts energy going out to the grid as well as energy drawn in. Finally the grid connection links the whole thing to the utility network, which acts as an near-infinite battery: it soaks up surplus and supplies any shortfall.

Running alongside all of this are two quieter systems: the mounting or integration that physically holds the modules (racking on a roof, or the substructure that makes a BIPV element weather-tight), and monitoring that reports how much the system is generating so faults show up. Modules, DC wiring, isolators, inverter, AC protection, consumer unit, meter, grid, mounting, monitoring - that is the anatomy. Learn it as a sequence and you can place, size and cost a system in your head before an engineer ever draws it.

The single most useful way to organise the whole system in your mind is to split it at the inverter into a DC side and an AC side, because they behave differently, carry different risks, and are governed differently.

Upstream of the inverter is the DC side: the modules, the DC cabling, the connectors, the DC isolator. Its defining feature is that it is energised whenever there is light - you cannot 'turn the modules off', only isolate their output - and it runs at a DC voltage that does not have the natural zero-crossings of AC, which makes DC arcs harder to extinguish and DC faults their own discipline. This is why the DC side has strict rules about cable type, routing, connectors, isolation and (increasingly) rapid-shutdown provisions, and why it is squarely engineering territory. For a designer the DC side is mostly about where the modules are and how their cables reach the inverter - runs that should be short, protected, and planned into the fabric, especially in a BIPV facade where cables must be concealed and still serviceable.

Downstream of the inverter is the AC side: AC protection, the consumer unit, the meter, the building's circuits, the grid connection. This is the familiar world of ordinary building electrics - the same alternating current your sockets already use - so it integrates into the building's normal distribution, but with added rules for connecting a generator: anti-islanding (the inverter must disconnect if the grid goes down, so it cannot back-feed a 'dead' line and endanger a lineworker), export limits, and metering.

The practical payoff of this mental split is judgement. When you hear that a system needs a 'string isolator on the roof' or 'rapid shutdown', that is the DC side. When you hear 'net meter', 'export limit' or 'anti-islanding', that is the AC side. The inverter is the border crossing between them. And the honest boundary of your role sits across this whole divide: the architect and interior designer shape *where the array and plant go and how they read*, but the sizing, protection, isolation and connection on both sides are binding engineering that belongs to the electrical engineer, the manufacturer's data and the codes - never to a designer's assumption.

A system that generates is only half the story; what happens to that power moment by moment shapes its whole value. At any instant the building has a demand (lights, cooling, equipment) and the array has a generation, and the two rarely match. Three things can happen.

When generation exceeds demand, the building uses what it needs and the surplus is exported to the grid. When demand exceeds generation - a cloudy hour, or the evening after sunset - the building uses all the solar it can and imports the shortfall from the grid. And in the sweet middle, all the generation is used on site as it is made: pure self-consumption. This matters commercially because, in most tariff structures, a unit you consume yourself is worth more than a unit you export: self-consumption offsets electricity you would have bought at the retail rate, while export is often credited at a lower rate (the exact rules are the utility's, and vary - Module 8.2). So two systems generating the identical number of units can deliver very different value depending on when they generate relative to when the building uses power.

This single idea explains much of what follows in the course. It is why load-matching matters (a building whose demand peaks at midday, like an office, is a natural fit for solar; a home that peaks in the evening is a poorer instantaneous match). It is why storage exists - a battery lets you park midday surplus and spend it in the evening, converting low-value export into high-value self-consumption (2.3). It is why orientation is not only about total yield: an array angled to catch late-afternoon sun may be worth more to an evening-peaking building than a higher-yielding noon-facing one. And it quietly shapes BIPV, where east and west facades generate at the ends of the day and can broaden the generation curve rather than piling it all at noon.

For the designer, the takeaway is to stop thinking of a solar system as a lump of annual kilowatt-hours and start thinking of it as a shape over the day that meets, or misses, the building's own shape of demand. The closer the two shapes sit, the more the system is worth - a theme we return to in Module 6. As always, the actual tariffs, export rules and net-metering terms are the utility's and the regulator's to set, not the designer's to promise.

Everything above is true of any PV system, rooftop BAPV included. But it lands with special force in BIPV, because integration changes where the parts go and what they cost.

First, the balance of system can dominate. On a simple roof, standard racking and short cable runs keep the non-module cost modest. In a BIPV facade, the 'mounting' is a bespoke weather-tight substructure, the cable runs are long and must be concealed yet serviceable, and the modules themselves are custom sizes and shapes. The module may be a smaller fraction of the total than a newcomer expects - which is exactly why BIPV must earn its place (Module 0). Seeing the full anatomy is what stops a designer from quoting 'just the panels' and being shocked by the real number.

Second, the plant needs a home. Inverters (and any batteries) need a ventilated, accessible, often cool space - a real architectural requirement. Put an inverter in a sealed, sun-baked cupboard and it will run hot, derate and fail early. This is a coordination task the designer owns: find the space, plan the cable routes, keep runs short, and give maintenance access, all early enough that it is designed in rather than bodged on.

Third, integration multiplies the DC-side considerations. A facade full of modules means DC cabling threaded through the building envelope, connectors buried behind cladding, isolation and (where required) rapid-shutdown provisions to think about with the engineer from the start. A module that is beautiful but unserviceable, or a cable route that cannot be inspected, is a design failure even if the array generates well on day one.

Fourth, the whole-system view is what makes you a good client of the engineer. You are not going to size the strings or specify the protection - and you should not. But if you understand the chain, you can ask the right questions (Where does the inverter live? How long are the DC runs? What is the balance-of-system cost? How does anti-islanding affect backup?), catch a quote that has hidden the plant space, and integrate the system honestly rather than treating it as a black box bolted to your building. The generation is the modules; the *system* is the anatomy - and design happens across the whole of it.