4 ms·

Has anyone done more detailed numbers on costing of roof mounted installation vs just putting utility scale solar on the ground? I would have thought if the ec

by jpgvm 4y ago

Has anyone done more detailed numbers on costing of roof mounted installation vs just putting utility scale solar on the ground? I would have thought if the economics were super favourable this would already be happening by now. To me it -seems- like a great idea but while I'm pretty versed in solar for home scale (diy off-grid setups etc) I have no idea how the numbers bear out for this install size, grid connection, etc.

algo_trader 4y ago
> Has anyone done more detailed numbers.. This is usually under "C&I" (commercial and industrial), as opposed to rooftop or utility scale. Try Lazard LCoE report, NREL Solar reports, IRENA renewables, or use google.
Syonyk 4y ago
> Has anyone done more detailed numbers on costing of roof mounted installation vs just putting utility scale solar on the ground? You're far cheaper on the ground, for a variety of reasons (at least in most areas). A good ground mount install should be able to come in around $1/W, and you'll have a hard time getting roof mount below closer to $1.50 or $2/W installed, and if you get $2/W, you're getting a great deal. If you're doing the work yourself, ground mount for around $1/W is doable (including the frames), but roof mount I've not been able to get below about $1.25/W, and that's hard - $1.50/W is more typical. The main difference is that the roof mount system requires rapid shutdown on (in NEC 2017 and later) every panel, and you're limited to 600V (though with rapid shutdown requirements, this is less relevant). For a ground mount system that's isolated away from random people (fenced area), you can run up to 1500VDC to the inverters, and this rather reduces your costs in wiring. You also don't need rapid shutdown, so you can just run strings of panels (20 or 25x 72 cell panels in series, depending on the environment). Also, if those strings have an arc fault somewhere, you're not going to burn up much that matters for a big ground mount install. You'll cook a couple panels, and that's about it. A fire under the panels on a warehouse is a much bigger deal. The main problem you run into with warehouses, though, is that there's just no good option for mounting. For a flat roof, they're generally built to exactly the loading requirements for the area. You don't have the spare PSF capacity for a ballasted mount, which means that assuming you've got the weight rating for the panels, you're looking at a lot of roof penetrations on a flat membrane roof to hold stuff down - and the odds of some of those leaking is basically 100%. Plus it's a royal pain to work on roof mount systems.
- jpgvm 4y ago
  Thanks for the great writeup! The root loading makes perfect sense when you think about it and in that context makes the equation much more expensive unless the warehouse could feasibly use all the power generated (which seems unlikely).
- s0rce 4y ago
  Do you know how the numbers compare for green field projects on the ground? They keep building new projects in otherwise wild/wilderness areas instead of on rooftops.
notatoad 4y ago
we massively under-value land in most cases, which is the reason that activists push for solutions like this. utility scale solar on the ground is economically better, if you're looking at "the numbers". but utility scale solar probably means repurposing farmland, because that's the cheapest land that has enough infrastructure built up around it to make a utility install easy. and repurposing farmland to build solar isn't a clear win in terms of overall societal benefit.
- Syonyk 4y ago
  > ... but utility scale solar probably means repurposing farmland... Great. Let's start by repurposing the land that grows the corn we turn into ethanol at somewhere around or below unity EROEI, as a political handout to the midwestern states. "Burning a megajoule of diesel, to grow enough corn to make something slightly less than a megajoule of gasoline," is not a winning solution to any problem but "How to best hand money to states that happen to have early Presidential primaries." It's not a bad gig for Iowa (I lived there for a decade), but neither is it a particularly good use of energy, fossil fuels, or money. Let's put solar up there, and work out. Once we've gotten rid of the ethanol debacle, we can see where things are and discuss from that point, but there is a lot of farmland right now that doesn't grow human or animal food, it grows corn that we process rather inefficiently into ethanol.
  - notatoad 4y ago
    i completely agree, but as long as this is all driven by economics the existence of those corn subsidies is going to mean the farmland that gets repurposed won't be the farmland that gets the most government subsidies, it'll be the farmland that grows the cheap food that people need to eat. i know ethanol farming is a waste, but it's a lot easier to convert a corn field back into a useful farm field before you've built a solar farm on top of it, rather than afterwards
    - shkkmo 4y ago
      > it's a lot easier to convert a corn field back into a useful farm field before you've built a solar farm on top of it, rather than afterwards Or you can do agrovoltaics and have both. There's some evidence that low density agrovoltaic setups increase corn yields.

by jpgvm 4y ago

algo_trader 4y ago

> Has anyone done more detailed numbers.. This is usually under "C&I" (commercial and industrial), as opposed to rooftop or utility scale. Try Lazard LCoE report, NREL Solar reports, IRENA renewables, or use google.

Syonyk 4y ago

> Has anyone done more detailed numbers on costing of roof mounted installation vs just putting utility scale solar on the ground? You're far cheaper on the ground, for a variety of reasons (at least in most areas). A good ground mount install should be able to come in around $1/W, and you'll have a hard time getting roof mount below closer to $1.50 or $2/W installed, and if you get $2/W, you're getting a great deal. If you're doing the work yourself, ground mount for around $1/W is doable (including the frames), but roof mount I've not been able to get below about $1.25/W, and that's hard - $1.50/W is more typical. The main difference is that the roof mount system requires rapid shutdown on (in NEC 2017 and later) every panel, and you're limited to 600V (though with rapid shutdown requirements, this is less relevant). For a ground mount system that's isolated away from random people (fenced area), you can run up to 1500VDC to the inverters, and this rather reduces your costs in wiring. You also don't need rapid shutdown, so you can just run strings of panels (20 or 25x 72 cell panels in series, depending on the environment). Also, if those strings have an arc fault somewhere, you're not going to burn up much that matters for a big ground mount install. You'll cook a couple panels, and that's about it. A fire under the panels on a warehouse is a much bigger deal. The main problem you run into with warehouses, though, is that there's just no good option for mounting. For a flat roof, they're generally built to exactly the loading requirements for the area. You don't have the spare PSF capacity for a ballasted mount, which means that assuming you've got the weight rating for the panels, you're looking at a lot of roof penetrations on a flat membrane roof to hold stuff down - and the odds of some of those leaking is basically 100%. Plus it's a royal pain to work on roof mount systems.

jpgvm 4y ago
Thanks for the great writeup! The root loading makes perfect sense when you think about it and in that context makes the equation much more expensive unless the warehouse could feasibly use all the power generated (which seems unlikely).
s0rce 4y ago
Do you know how the numbers compare for green field projects on the ground? They keep building new projects in otherwise wild/wilderness areas instead of on rooftops.

notatoad 4y ago

we massively under-value land in most cases, which is the reason that activists push for solutions like this. utility scale solar on the ground is economically better, if you're looking at "the numbers". but utility scale solar probably means repurposing farmland, because that's the cheapest land that has enough infrastructure built up around it to make a utility install easy. and repurposing farmland to build solar isn't a clear win in terms of overall societal benefit.

Syonyk 4y ago
> ... but utility scale solar probably means repurposing farmland... Great. Let's start by repurposing the land that grows the corn we turn into ethanol at somewhere around or below unity EROEI, as a political handout to the midwestern states. "Burning a megajoule of diesel, to grow enough corn to make something slightly less than a megajoule of gasoline," is not a winning solution to any problem but "How to best hand money to states that happen to have early Presidential primaries." It's not a bad gig for Iowa (I lived there for a decade), but neither is it a particularly good use of energy, fossil fuels, or money. Let's put solar up there, and work out. Once we've gotten rid of the ethanol debacle, we can see where things are and discuss from that point, but there is a lot of farmland right now that doesn't grow human or animal food, it grows corn that we process rather inefficiently into ethanol.
- notatoad 4y ago
  i completely agree, but as long as this is all driven by economics the existence of those corn subsidies is going to mean the farmland that gets repurposed won't be the farmland that gets the most government subsidies, it'll be the farmland that grows the cheap food that people need to eat. i know ethanol farming is a waste, but it's a lot easier to convert a corn field back into a useful farm field before you've built a solar farm on top of it, rather than afterwards
  - shkkmo 4y ago
    > it's a lot easier to convert a corn field back into a useful farm field before you've built a solar farm on top of it, rather than afterwards Or you can do agrovoltaics and have both. There's some evidence that low density agrovoltaic setups increase corn yields.