small scale solar like this is quite inefficient compared to grid-scale stations.
grid-scale generation is much more effective, both in cost and generation efficiency.
Can you explain a bit better on the magnitude of these differences and why they exist? Thx
Solar panels are most efficient when sunlight hits them as close to perpendicular as possible. Large scale arrays tend to be 1 axis sun tracking, which yeilds about twenty percent more power with the same panels compared to optimal angle fixed arrays at cost of some small motors, a computer, and incresed space between the panels. Small scale and or rooftop by contrast tends to be space constrained, and often ends up at bad angles.
It still works obviously, but between the fixed angle and cramped space you see significant losses compared to what the same pannel could output in a large scale array.
Cost wise, the most expensive part of any small scale solar array tends to be the labor to install it. Panels are now cheaper per square m than fence posts, inverters are expensive but not that expensive, but design and construction are not. It is a lot faster and easier to have a team start at one end of a field and put down racks in assembly line fashion than having someone come out and design a system, another team climb roofs to install conduit and brackets, then a bunch of electrical work, before finally getting an electrical inspector to come out and sign off on it.
None of this is to say that small scale is useless or you shouldn’t do it if you can, especially in California where Pacific Gas and Energy have spent the last forty years outright refusing to do any maintenance or infrastructure investments unless the govement picks up the entire tab and lets them raise some of the highest rates in the nation, but there definitely is an argument to be made that if the government is paying for it either way than spending should go to the place that gives an extra twenty to thirty percent output for the same cost.
Good points.
So this leaves us comparing the inefficiency of fixed-angle panels to the inefficiency of transmission over a grid. IIUC, grid inefficiency is huge. So wouldn’t it be wise to upgrade the home kits to add sun tracking and a motor?
Very few homeowners are going to want to pay even more for tracking. You’d also need more space to accommodate the swing of the array, and it also brings about more consideration for wind loads. Grid inefficiencies are still present with rooftop systems as well, as excess power is sent back to the grid.
The US national average grid loss is about 5%, which I figured would often be more than made up for by a lack of shadeing at larger installations. While grid losses are likely to increase as we move to renewables, that’s down to the fact that prime solar and wind spots tend to be much farther away from major north American cities than the suburban power plants that feed them currently. Given the density of said cities, rooftop solar just takes to much space to provide the necessary power for a taller building.
While you can find some home scale ones with sun tracking, given the weight and foundation requirements they almost always need to be built on the ground, and do cost even more. You also have the aesthetic constraint, as while panels flat against the roof only annoy the most Nimby of Nimbys, large person sized poles like this are harder to hide, even if i think they look neat crystal flowers.
Finally, tracking panels do have higher routine maintenance requirements, as there are now parfs that can move, wear, and jam. It doesn’t happen often, so one technician can dozens of square miles by themselves, but it takes a lot more time when you need to schedule around individual homeowners and a wide array of different types of install as compared to see a notification on the monitoring software when you get into work, hop in truck with a set of standerd replacement parts, go to panel and fix.
A couple things to add to sonori’s reply, the system efficiency also goes up when you can use higher capacity inverters. Essentially, one grid-scale facility (say, 100 MW) will have far fewer inverters than 100 MW of rooftop systems, meaning less efficiency loss.
While larger facilities can be optimized for azimuth and sun angle or with tracking, rooftops are often not optimally aligned, leading to a decrease in capacity. Sure, you could add tracking to your residential system, but that’s a significant cost that many homeowners won’t see the benefit of. There might also be obstacles that shade a residential system, which is often addressed when selecting sites for larger systems.
I’m not disagreeing with anything you’ve said, but there is a bit of nuance on optimal angle. I’m also in SK (about 50.7 degrees), and when I was looking at putting up panels, it was cheaper to throw a couple of extra panels up than to modify the install for “optimal fixed angle”.
Obviously, starting from scratch, you’d be foolish to not include “optimal fixed angle” or maybe even a manually operated dual-angle system that has a “summer position” and a “winter position”. For my install, a dual-angle install would have required the same number of panels as an optimal fixed angle. I would have got a larger margin of overproduction, but not by enough to eliminate a panel.
A lot of what I see on solar panels is still stuck using the same analyses as when they were less efficient, more sensitive to clouds, and more expensive. It wouldn’t surprise me to learn that even large scale installations would be better served by eliminating the expense and maintenance costs of tracking instead of pushing for maximum performance.
(Note: we didn’t ever get the install, because we couldn’t get financing, even when SaskPower was still doing full retail buyback.)
Yeah, standard installations will often be cheaper than a customized mounting. I’m at 52.9 and have a rooftop system with two arrays directed SW and SE because of rooflines (sized so the house is net-zero - 39 panels for 14 kW which gets me about 15,000 kWh annually), but I’d definitely go ground mount with a summer/winter position rack if I were to do it again. I work a bit with simplified PV modelling now and am disappointed that I didn’t know more sooner, probably would’ve saved quite a few dollars.
We also did an early feasibility study for a 10 MW farm and did find that tracking was marginally better, even including maintenance costs, but that does introduce the risk of labour availability and cost fluctuations that isn’t there on fixed mount. Overall, the project wasn’t feasible with the rates SaskPower pays.