[Bored chemist]

Only if its focal distance is 7.5 x 1013m, giving its radius of curvature of 1.5 x 1014m - what most engineers would call "flat within spec".

The spec would be " a sphere".

[Bored chemist]

Note that  a practical tunnel has an Ω cross-section, not a circle - that is called a pipe!

Best tell the people who bore tunnels.

They think that circular cross sections are wonderful.

[wolfekeeper]

Even the best solar panel reflects some of the incoming radiation. The curved office building at 20 Fenchurch Street has caused serious problems

Those aren't solar panels, it's just window glass, mounted almost vertically along a curve. In the UK, and similar latitudes, solar panels are nearly always angled at less than 40 degrees to the ground, which reflects the sun back into the sky, nor is there any reason to architecturally arrange them to cause focusing.

See:

https://blog.spiritenergy.co.uk/contractor/best-angle-solar-panels-uk

[alancalverd]

Best tell the people who bore tunnels.

They think that circular cross sections are wonderful.

Best tell the guys that built the one in your picture! Modern boring machines like to make circular tubes because rotation is easy to generate and control, but you then have to fill or cut out the bottom to accommodate the flat road - and even the track that supports the boring machine. Hand-dug tunnels like the one shown can have an omega section right from the start, so the hydrostatic forces compress the brick lining but the diggers are always working from a flat surface.

[alancalverd]

Only if its focal distance is 7.5 x 1013m, giving its radius of curvature of 1.5 x 1014m - what most engineers would call "flat within spec".

The spec would be " a sphere".

Are you seriously suggesting that the earth is not flat? Begone, heretic!

[Bored chemist]

Rather mundanely, I'm suggesting that a Dyson sphere is a sphere- though , technically, other shapes could work, with less efficiency.

Best tell the people who bore tunnels.

They think that circular cross sections are wonderful.

Best tell the guys that built the one in your picture! Modern boring machines like to make circular tubes because rotation is easy to generate and control, but you then have to fill or cut out the bottom to accommodate the flat road - and even the track that supports the boring machine. Hand-dug tunnels like the one shown can have an omega section right from the start, so the hydrostatic forces compress the brick lining but the diggers are always working from a flat surface.

Well Damn it.
Why didn't someone mention the flat bit before.

Oh, I did.

, since he's talking about a curved surface, we can ignore the floor of the tunnel.

[TommyJ]

If you can move the receptor, then a flat plate can always be 100% covered at optimum angle (i.e. perpendicular to the sun). If you aren't going to move the plate, aligning it to the average midday sun will give you the maximum yield for a given area of cells, over the year. 

Also, the rotation would be effective, if your roof (or any receiving area) is some acres. Other wise each plate would hardly need any rotation. And automation cost will be significant compared to the outcome.

[ambertcarrero]

Is that the tunnel that you could go time travel or in the other parts of the galaxy?

[stevesmith6718]

The proposal is inefficient because not all the elements will be optimally irradiated at any time.

The daily arc of the sun varies (from none to 360 degrees at the arctic circle!) depending on the time of year. The maximum elevation varies with season.  So you need to adjust the alignment of any receptor to maximise its yield.

If you can move the receptor, then a flat plate can always be 100% covered at optimum angle (i.e. perpendicular to the sun). If you aren't going to move the plate, aligning it to the average midday sun will give you the maximum yield for a given area of cells, over the year. 

Depending on the use of the electricity, it may be better to angle the plate a little lower, to maximise winter yield.

Thanks for the suggestion

[wolfekeeper]

I've looked into this a bit further. Angling the solar panels so that they point directly at the sun at all times gives the maximum electricity for any given amount of solar panel, and fairly flat power output from sunrise to sunset. However the mechanism to do it is expensive.

But apparently around 2014 the solar panels themselves had dropped in price enough that it became cheaper to simply use more fixed solar panels oriented at different angles and lose some of their potential production, It's actually more economic to do that in many cases than just using a south facing panel, particularly if you're using the power directly rather than exporting it (where you'd want to be replacing electricity at ~20p/kWh with solar power at ~6p/kWh.)

For example you could point them south west and south east at about 90 degrees to each other and you'll get pretty much the same effect as a solar tracker. Or East and West facing sloping roofs work really well in summer (winter sucks though a bit unless you add a south facing panel as well.)

There's a few more interesting wrinkles as well that make it particularly worthwhile in some cases.

If the panels (or strings of panels) are the same voltage, then you can parallel them up and connect them to a single inverter and the power will add. Obviously you can't get more power than the inverter can handle, but the inverter won't take more current than it can use, so it's safe and you'll get more power than normal when it's cloudy because the light scatters and gets absorbed by the more shaded panel as well.

If you have a flat roof or the panels are on the ground, then you can manually re-angle the panels differently a couple of times a year. That really improves the energy output, maybe 35% extra kWh per year, mostly in the mornings and evenings and particularly in winter, because you can point the panels south.

It turns out that the extra panels are relatively cheap, the panels and frames are only about 20% of the cost of the system or something so that's why it's a net win (at least sometimes particularly when you're self consuming the electricity and want a relatively even output curve). Of course if you're looking for total maximum output kWhs and you're only being paid per kWh and there's no limits on power output, then pointing all the panels due south may be superior, but then you'll have to scale the inverters and everything else.

Nevertheless some solar farms now do this, particularly if they're limited by their grid connection (which is often the case) and so they want a relatively flat power output.

[JesWade21]

If you're interested in the field of sciences of materials, you must seriously consider the perovskite solar cell (PSCs)! They're a cool and rapid-moving technology. Since the beginning of their development in 2009, they've been able to increase their efficiency from 3% and up to 22% of efficiency which makes the technology one of the fastest-growing technologies available today. The most interesting thing about perovskites (and the reason they wrap into tandem cells so well) is that you are able to "tune" the band gap of the absorption layer in many different ranges dependent on the quantity of bromide or iodide present in the mix. They're also semitransparent and behave like an optical splitter. This makes it possible to create customized tandem cells based upon the characteristics of your "bottom layer" absorber (oftentimes silicon wafers, though other organic cells have also been utilized).

PSCs are easy to create, however they can be difficult to master since they can be sprayed with a coating on the glass, or on any surface with electrodes and voila! You've created a solar cell (see transparent solar panels for sky scrapers, super amazing technological advancement!). The issue is that they have many stability issues in PSCs that are present in the present and require addressing before they can be marketed as an actual product for sale however, given the rapid technological advancements, I believe we'll be there in the next 10 years!

[Rocky6419]

Hi, everyone here I can tell why are solar panels not curved?

 Because most of the radiations are captured by the plane surface, solar does not need to have a curved shape.

[Kartazion]

Hi, everyone here I can tell why are solar panels not curved?

Curved the mirror will concentrate all the energy in one point. No photovoltaic cell could withstand this intensity. For proof you have the solar oven of Odeillo

Odeillo solar furnace 

[Kartazion]

[ramisthand76]

Solar panels do not need to have a curved shape. Maximum radiation is captured from a plane surface. Curved shapes as seen in solar thermal plants are used to focus the radiation to a single point, thus magnifying the local energy density. Focusing is not necessary in PV, therefore you don't get to see curved panels.

[mayorovd]

You want all cells in the solar panel to be similarly illuminated, so they put out a similar voltage, have a similarly low resistance, and the inverter can efficiently convert the total DC voltage (whatever it adds up to) into AC.

[wolfekeeper]

In operation, solar panels are more or less constant voltage, largely independent of illumination. The current depends on the illumination. Because of that, panels with different illuminations work well in parallel, less so in series. You want panels in series to all have similar illuminations, to all be at the same angle to the sun.

[evan_au]

You want panels in series to all have similar illuminations, to all be at the same angle to the sun.

Half of my panels face east, half face west.
- One side active in the morning
- One side in the afternoon
- both semi-active at noon
- The inverter has two inputs, one for each string of solar panels.

The ideal would be to have them all facing north, but my house doesn't have a large north-facing area - and there is apparently some local regulation that discourages solar panels on the road-facing side of the house (but I'm sure I've seen other homes doing it...)

[wolfekeeper]

The best angles to have your solar panels at is complex. Since about 2014 the cost of the panels are low enough that having more solar panel capacity than inverter capacity became cost effective. I believe I'm correct in saying that you would ideally want your panels facing north-east and north-west in your case; the cost of the extra panels is a small proportion of the cost of the system. When I ran the numbers on a system I was planning here, it was something like the cost went up by 15% and the productivity went up by 20%.

East-west systems approximate solar tracker outputs and may be optimal near the equator, and if you're not on net metering where the feed in tariff is small compared to the normal usage tariff if you're doing any significant self usage of the electricity.