Installation
The gap behind the panels is doing more work than it looks
Air moving through the cavity between modules and roof carries away heat that would otherwise be paid for through the temperature coefficient.
By Manish Trivedi4 min read

Panels convert a minority of what lands on them
Most of the energy arriving at a module isn’t converted to electricity. Some is reflected, a share becomes current, and the remainder becomes heat in the laminate. That heat has to leave, and the only routes available are radiation from both faces and convection into the air around them.
The reason it matters is the temperature coefficient. A crystalline module produces less as it warms, at a rate given on its datasheet, so every degree the cell sits above ambient is a small, permanent loss for as long as the sun shines. Across a summer that adds up to a quantity worth caring about.
Nothing about the module changes this. The mounting does, because the mounting decides how easily the heat gets away from the back face.
A cavity works by buoyancy, and buoyancy needs two openings
Air warmed by the rear of a module becomes less dense and rises. If it can escape at the top of the array and be replaced by cooler air entering at the bottom, a slow continuous flow establishes itself and carries heat away without any fan or moving part. That is the whole mechanism, and it is remarkably effective for something entirely passive.
It stops working if either opening is blocked. A cavity open at the bottom and closed at the top fills with hot air and stagnates. The same is true in reverse. This is why bird protection mesh around the perimeter of an array, which is a sensible thing to fit where pigeons are a problem, should be a mesh with real open area rather than a solid barrier.
Depth matters less than people expect once a reasonable gap exists. The standard rail height on a typical on-roof mounting system is adequate, and increasing it further gives diminishing returns while adding wind loading and visual bulk.
In-roof systems trade cooling for appearance
Integrated systems replace the roof covering across the array area, so the modules sit within the roof plane rather than above it. They look considerably better, they remove the standard bracket penetrations, and they are the natural choice when a roof is being re-covered anyway.
The cost is thermal. Sitting the modules in a tray, close to the roof structure, restricts the cavity and reduces the airflow behind them, so an integrated array typically runs hotter than an on-roof one in the same location. The yield penalty is real, usually modest, and worth knowing about rather than discovering later.
There is a second consideration that has nothing to do with electricity. The roof is now partly a manufactured system rather than overlapping tiles, and the detailing around it has to manage water and, in some constructions, moisture from inside the building. Where the trade knows the product, this is routine. Where it does not, it is where the problems come from.
Flat roofs and ground mounts get this for free
A ballasted array on a flat roof sits above the surface on frames with open air on all sides, so ventilation is rarely the issue. What can compromise it is a wind deflector fitted across the back of each row, which does useful work reducing uplift and can restrict the airflow that would otherwise pass through.
Ground-mounted arrays are the best case of all, with unrestricted air on both faces and no hot roof surface radiating from below. They routinely run cooler than roof arrays in the same climate, which is one of the quieter reasons their yield per unit of capacity tends to be higher.
In hot climates this difference stops being marginal. Where ambient temperatures are high for months, the gap between a well-ventilated array and a poorly ventilated one is one of the larger avoidable losses in the whole system.
What the roof underneath gets out of it
Shading a roof does change what happens beneath it. The covering under an array absorbs far less solar radiation than the exposed covering next to it, so the loft below runs slightly cooler in summer, which is a small comfort benefit in a hot climate and irrelevant in a cold one.
That shading also slows the weathering of the covering itself, since ultraviolet exposure and thermal cycling are what age most roofing materials. It is a genuine benefit and it isn’t a reason to install solar, since the fixings introduce their own vulnerability at the same time.
The practical rule is short enough to state in a sentence. Leave the cavity open at the top and the bottom, don’t stuff insulation or bird boarding into it, and do not lay modules directly against a deck unless the system was specifically designed to work that way. Everything else about ventilation follows from that.
Common questions
Does a hot loft mean my panels are running hot?
The two are related and not the same thing, since the loft temperature reflects what the covering absorbs and passes inward while the module temperature reflects what the cavity can carry away. A hot loft does suggest a roof absorbing a great deal of heat, which raises the temperature of the air the array is trying to cool itself with.
Should I fit bird mesh around the array?
Where pigeons nest under panels it is worth doing, because the debris and droppings that accumulate cause soiling, blockages and occasionally damage to cables. Choose a mesh with genuine open area and fit it so the cavity can still breathe at the eaves and at the ridge, rather than sealing the perimeter solid.
Do cooling systems for panels work?
Water cooling and similar schemes do reduce cell temperature and do raise output, and for a domestic rooftop they add pumps, plumbing, maintenance and a parasitic load to recover a few per cent. It is a good demonstration of the physics and a poor purchase for a house, which is a combination that describes several ideas in this field.
Deputy editor, Power Your Roof
Manish has been reporting on solar basics, batteries, bills & tariffs since long before it was fashionable and would rather show the working than assert the conclusion.





