Keeping those panels cool

The people from Caname, the roof makers, and Rooftech, the roofers, came to visit the other day. My concern was with the air channel under the panels, which seems to me to be just too small. I've started measuring the temperature of the air coming out of the top, and it was getting up to 70 degrees centigrade. You can see a graph of the temperatures below, showing also for references the ambient temperature, the temperature inside the house and the temperature at the bottom of the slab, which comes pretty much to a straight line. The heat of the actual panels is going to be more than the temperature of air in the channel. 

So what? I hear you ask. 

I can hear some of you replying that solar panels produce less electricity as they get hotter. With the Suntech STP190S-24/Ad+ panels we are using, the efficiency drops half a percent with each one degree increase in temperature. 

As a thermodynamic system, incoming heat is beating down in solar radiation. This heat is lost in four ways: the panels are losing some heat to the wind from the top of the panels by convection, they are losing some heat by convection to air passing through the channel between the panels and the roof, they are losing some heat that is converted to electricity and they are losing some through the top by radiation. A small amount of heat will be conducted from the panels to the roof, but the roof is well insulated, so this heat is not really going to go anywhere very quickly. Heat that is not lost will make the panels hotter, and the efficiency will go down.

Directly fitting panels onto roofs as solar tiles is a bad idea because of this heat loss. Even conventional arrays that have been added onto a roof with a gap underneath will suffer efficiency loss of around 20% in the summer. 

Caname have done some research with panels fitted directly onto a roof with no air channel, some fitted in the conventional way, where air can flow north to south and east to west, and some fitted in their roof system where air only flows south to north. Their results over a year peg the conventional roof at 100%, find that fitting the panels directly to the roof with no gap drops to around 80%, while their roof is 99.5%. In terms of average panel temperatures, the conventional panels and their panels averaged around 65 degrees, while the tile style was over 70.   

In my opinion, rather than treating the conventional way as 100%, they should be treating ideal output as 100%, and ideal output means either full rating of the panels, or the projected power output at ambient temperature, in other words with perfect cooling.

The fact that they are building a roof system should mean that they can do better than panels that are added to an existing roof. Setting this as their target seems to be aiming too low. For example, they mount the panels on corrugated steel. Corrugated steel roofing may sound like a really bad idea, but it should work very effectively to cool the panels, both by channeling the air from bottom to top of the roof, and by increasing the surface area to conducting the heat from the panels.