It is not true that hexagonal leather pieces are the best geometry to make a spherical soccer ball. Most modern footballs consist of twelve regular pentagonal and twenty regular hexagonal panels positioned in a truncated icosahedron spherical geometry.
Of course, most roofs are not spherical so hexagons would work. In fact the solution to your problem might be a combination of hexagons and half-hexagons or triangles.
The optimum fill shape for a solar panel will be the size of a single cell, created from a mono-crystalline silicon ingot. I think right now they are about ?? 10" across?
They need to be connected in series to get an efficient voltage for either your bulk inverter or your roof-top micro inverters. The cost and difficulty of mounting individual cells is worse than the problem of interconnecting them, and the low cell voltage makes per-cell inverters inefficient even if they were economic.
People who actually want to fill roofs use poly-crystalline cells, which can be cut to match the desired shingle-size. I don’t know how they are interconnected, but they are mounted like roof tiles. I don’t think they are using micro-inverters at all, because the cell performance is not comparable to the performance of a large mono-crystalline panel.
Traditional panels are defined by three things: the matching characteristic of a lead-acid battery, the size of the silicon ingot the cell were cut from, and the size and shape of the kind of panel people wanted to manufacture and handle. 32 cells gave about the right voltage, could be mounted in a 4x8 pattern, in a panel small enough to handle but big enough for economy.
Perhaps the better alternative is to start building homes with solar as part of a total equation of eco design. Don’t put roofs with odd shapes and try to have a good southernly aspect with the right pitch for that area. Yes it’s a longer term solution, but it seems like a sounder option then what you suggest.
But back to your inquiry, they already have solar roof tiles but from what I heard their lifespan is not been the greatest. Perhaps we could make the panels very long and skinny of varying lengths, or maybe can be cut on site to fit, perhaps with certain indicator lines indicating where cuts can be made.
Each panel also has an inert perimeter. A smaller panel will have a higher perimeter-to-area ratio; make your panel small enough, and the actual area of photovoltaic power output would fall to zero. So bigger panels should be more efficient in terms of power per installed area. But if your panels are too big, they’re a pain to transport and install, and you do in fact end up with a bunch of wasted perimeter on typical household roofs.
What are the odds that current panel manufacturers are way off of the mark as far as optimal panel size and geometry?
It wouldn’t surprise me if solar panel mfrs have done exactly this sort of analysis. It also wouldn’t surprise me to learn that such analyses were proprietary and confidential.
And I’d be surprised if anyone but the panel manufacturers had ever done such a study. It’d be a hard study to do, and hence expensive, and nobody else would have the financial incentive to fund it.
Yes, this has to be considered in the real world. Local building inspectors have been dealing with the local electric company for years, and are real friendly. They tend to put whatever barriers they can against solar panels.
When my solar panels were installed, I expected resistance & delay & silly fees from the electric company (I’d been warned of that), but I was surprised by the passive opposition & added ‘requirements’ by the City building inspector office. Maybe it’s wishful thinking, but you always seem to feel that corrupt government officials are somewhere else, not in our town!
Traditional panel sizes are traditional. They are based on a time when solar panels were far more expensive than they are now.
Part of the reason why solar panels are so cheap now is that solar panels are produced by hundreds of Chinese companies that got into the business in a typical Chinese-economy boom-bust bubble. Those companies weren’t doing research on the optimum panel size – they were just trying to get into a business where it was apparent that demand was growing and that a good margin was available.
But panel size and geometry isn’t set by manufacturer research anyway. It’s set by installers. Right now installers are using bigger panels, because they’re easier to install and work well with bulk inverters.
If the world actually moves to micro-inverters (and the world hasn’t done that yet), we may see another change. Or we could move to roof-replacement, instead of on-roof panels.
There are a lot of other constraints involved in designing a useful solar array. Covering the roof seems to assume the panels lay flat on the roof, no matter what the pitch of the roof. This leads to all manner of inefficiencies. Putting any panels on the shaded side of a pitched roof is simply wasting money. But other matters come into play.
The average Joe has a day job. So power generated during weekdays isn’t going to be used in the household. You either sell it at whatever feed-in rate you can get - which can vary wildly depending on government policy/subsidy and your supplier, or you buy a battery, with the significant costs that comes with. So you may want to optimise panel layout and angles so that they optimise efficiency when the power is most useful to you. This starts to drive the layout towards both optimal elevation angles, and and a mix of azimuths. If you have a flat roof the one thing you will never do is lay the panels flat. If you have a pitched roof your ability to optimise the angles is more limited, but optimise you will. Simply coating the roof in panels flat to the roof line will not provide an optimal solution, even if you can cover the entire roof.
Inverters are going to get more interesting. I’m pretty sure we will see a major swing to inverters built to provide “inertia” to the grid, rather than slaving themselves to the grid. The technology already exists and there have been real world trials. It makes for more complex inverter technology, but solves some significant problems when the grid starts to move to dominant renewables.
These extra fees, and the labor and liability with roof installation hugely increase the cost of solar panels.
Hugecost increases. Per the source, solar done at an individual rooftop level is $81-$170 per megawatt-hour but only $36-$44 done at utility scale.
This is because the utility company can get the necessary permits/do the engineering work for a huge, square kilometer or larger plot of land. Where the plot is somewhere the land is cheap and flat. So about the same amount of work in terms of planning and permits as it takes for one house, and then they cover a square kilometer with panels instead. Also they get bulk discounts on the hardware, and it’s on flat land, so the labor and installation is far quicker and cheaper. (an electrician, rather than coming to one house at a time, gets paid to walk down entire rows of the array doing the connections and inspections)
Basically, rooftop solar doesn’t make much sense. Perhaps if it was a plug and play product*, where Federal law forces the locals to accept products made to a certain standard, and it was only on new houses/during new roof installs it would make sense.
(*this is a thing. Consider this - you buy a microwave oven from costco. Feds and UL approve it. Your local jurisdiction doesn’t get a say whether you can plug that oven in)
As for panel geometry: I went here and choose an area in Clear Lake City Texas, where I used to live.
I picked a house similar to the one I used to live in. Using the calculator, the South facing part of the roof can support 100 square meters of panels, easily, with some room leftover. This is 15 kW DC of panels with the cheapest kind available, at 15% efficient.
These houses (2600 square feet, hot climate) consume about 1000-1500 kWh per month. (more if occupied by a full family and using electric hot water heating)
This means that every month, the panels would be producing 1821 kWh but the home only consumes 1000-1500.
So there is not a *need *for some fancy technology to squeeze more performance out of the roof area.
But if you use better panels - “premium” on PVwatts is only 19% efficient but you can routinely buy 22% efficient panels for a modest cost increase - you can produce 2783 kWh per month. If both adults work, and have 60 mile commutes, and both drive Tesla model 3s that get 240 watt-hours per mile, and are charging at 93.3% efficiency.
Then the household would be using 616 kWh to “fuel” their vehicles and thus need to produce ~1500-2100 kWh per month to be “net zero”.
Optimal sizing then depends on local electric company policies, the cost of the solar installation, and so on.
In reality while it’s cool to do these numbers, the most economically efficient thing to do, in many career fields, is to be flexible and move often for better compensation. So solar panels and other fancy upgrades don’t make any sense, you won’t live in a place long enough to benefit.
Instead what makes sense is to rent a place to live, and just deal with the extra cost of the utilities, they end up being small potatoes. (compared to how much buying/selling a house or apartment rent is)
For the above: I used PVwatts, entered the address, and drew on the satellite picture of the roof. PVWatts Calculator
Try it for your area, then find out what the annual solar production for your area actually is. I bet you’ll be surprised: if you live in a single family home, and are in the continental USA, no matter where, you can probably generate all of the electricity the house needs with just the solar on the South facing side.
Some folks might need to upgrade their home to reduce consumption: upgrades include:
a. Go from 12-15 SEER (central) to 39 SEER (so 1/3 the power used) mini-split air conditioners. (Gree or other brands)
b. Upgrade to on-demand electric hot water. (or on-demand gas if you live in the North and have natural gas available)
c. All LED lighting
d. Energy star appliances
e. More recent TVs and computers (huge decreases in power consumption in recent years)
f. Cellulose spray the attic for a thicker insulation layer
g. Seal air leaks to the outside
Note that for me, just doing (a) made an enormous difference. I was using 1500 kWh in the summer and this dropped it down to 600 or less. There was no need to bother with sealing air leaks or better insulation because I had simply made cold/hot air so cheap it didn’t matter.
The problem here is that a square kilometre of solar farm can’t be used for anything else, the land value disappears. With solar on a roof top, I can have a perfectly useful house holding it off the ground. There is no other value to sloped roof area.
This would be a valid comment except there’s large tracts of land that aren’t doing much right now and the land is nearly worthless.
https://www.bloomberg.com/graphics/2018-us-land-use/
Basically just grazing land, and I think it’s possible to keep using land with solar panels as grazing land (since water is the rate limiting factor, not sunlight) with the panels on poles.
Then invest in a solar fence. Yep, a fence made of solar panels. Lots easier to install, I expect.
There is a company (German, I think) that will put solar panels above farmland, so you can still grow crops under them. The panels are on a framework that’s like 3 or 4 meters above ground, high enough so a tractor can be driven underneath. You might wonder about how the plants get enough light. Well, the solar panels have enough gaps between them to let some light down. Plants don’t use anywhere near 100% of sunlight, more like about 2%. And some crops will actually grow better in those conditions. Also, having plants underneath keeps the panels cooler, so they don’t lose efficiency when it’s gets hot. I’d give a link, but I don’t remember the name of the company.
Fair enough. Thanks for that Bloomberg link, very interesting info graphic!
I’ve driven past houses with solar panels on their roof. I don’t know the size but they’re large panels - 4’x6’ or 4’x8’ probably. Then there’s one huge gap because there’s something on the roof, typically the bathroom exhaust stack where they don’t put any panel in that area. It seems to me that having a half-sized panel over whatever the standard size is could be used to fill in that gap as, percentage-wise, missing one large panel on a roof is a big hit. Why don’t I see that? Is having two standard sizes so difficult to make/keep in inventory?
It’s most likely because panels need to be matched pretty closely in terms of voltage and current output.
According to this Zillow study, Homes With Solar Panels Sell for 4.1% More.
So with a little luck the next guy will pay some or all of the solar panel investment.
Fair enough. It depends on how long you intend to live in one place. (And, really, your expected value of residency). We live in a world where your best guarantee of career stability is to have in demand portable skills such that you can get another position elsewhere easily. Loyalty to a single company is just being a victim.
That’s correct, if you do not have microinverters (one inverter dedicated to each panel; described above) all the panels have to be identical. The output on a traditional system will only be as good as the worst panel.
Additionally, there are clearance rules in the code about how close a panel can some to specific roof features. I can’t say specifically how those apply to penetrations like exhaust stacks, but I know the layout on my roof was constrained by several things, including distance to the ridgeline, distance to a chimney, and other things like that. So it could be even if the install was microinverter based, a part-sized panel (assuming they exist) couldn’t even be deployed in that spot per regulatory stuff.