Tweet
I'd suggest any frame modification might not be worth the effort. Structural considerations dealing with panel rigidity are more important than heat transfer considerations from improved air flow under a panel. Having at least 6" of clearance under a panel is probably more important, easier to achieve and more practical.
I'm not knocking your efforts, but unless you've got a lot of test equipment to monitor before/after, side/side performance, I'd be careful about inferring too much. There's not a whole lot that's new in the heat transfer business, with most any practical and workable improvements to lower PV operating temps. and thus increase efficiency having been tried. Most large arrays and all residential ones I know of or have heard about do not use aux. cooling or radically modified designs to lower panels temps.
My experience is that most schemes to improve efficiency via lower temps. come mostly from people who know little about solar energy engineering and even less about heat transfer.
While not discouraging your efforts or enthusiasm, if there was a better, more practical and most importantly cost effective way to do it, it would have happened by now.
As you and others note, the trick is to do it effectively, meaning safe, workable and cheap. We ain't there yet.
As for natural convection, while important both above and below a panel mounted parallel to a roof, it will be quickly overcome by any forced convection of even a moderate wind, and is usually close to or about an order of magnitude less than the cooling that even a light breeze of about 1 m/sec. might cause. The obstructions under an array, standoffs, flashings, cabling, etc., may even enhance the heat transfer rate under an array by increasing turbulence, depending on the application and dimensional particulars.
I'd suggest checking the open literature for information about natural (gravity induced) and forced (wind driven) convection and how it affects panel and array temps. Since my data for my location and array is more voluminous than can be transmitted here, and is mostly a confirmation of what's already in the open literature, but somewhat specific to my particular application, and also more extensive than what can be transmitted here, the answer to your last question is a respectful decline. See a decent undergraduate text on heat transfer, then see Duffie & Beckman. You'll get more out of the time spent.
Without such information, and the background it takes to understand what's contained in those sources you're mostly spinning your wheels and/or covering ground already travelled anyway.
I'm not knocking your efforts, but unless you've got a lot of test equipment to monitor before/after, side/side performance, I'd be careful about inferring too much. There's not a whole lot that's new in the heat transfer business, with most any practical and workable improvements to lower PV operating temps. and thus increase efficiency having been tried. Most large arrays and all residential ones I know of or have heard about do not use aux. cooling or radically modified designs to lower panels temps.
My experience is that most schemes to improve efficiency via lower temps. come mostly from people who know little about solar energy engineering and even less about heat transfer.
While not discouraging your efforts or enthusiasm, if there was a better, more practical and most importantly cost effective way to do it, it would have happened by now.
As you and others note, the trick is to do it effectively, meaning safe, workable and cheap. We ain't there yet.
As for natural convection, while important both above and below a panel mounted parallel to a roof, it will be quickly overcome by any forced convection of even a moderate wind, and is usually close to or about an order of magnitude less than the cooling that even a light breeze of about 1 m/sec. might cause. The obstructions under an array, standoffs, flashings, cabling, etc., may even enhance the heat transfer rate under an array by increasing turbulence, depending on the application and dimensional particulars.
I'd suggest checking the open literature for information about natural (gravity induced) and forced (wind driven) convection and how it affects panel and array temps. Since my data for my location and array is more voluminous than can be transmitted here, and is mostly a confirmation of what's already in the open literature, but somewhat specific to my particular application, and also more extensive than what can be transmitted here, the answer to your last question is a respectful decline. See a decent undergraduate text on heat transfer, then see Duffie & Beckman. You'll get more out of the time spent.
Without such information, and the background it takes to understand what's contained in those sources you're mostly spinning your wheels and/or covering ground already travelled anyway.
Comment