6 ms·

If it's not off-topic, I'm curious to learn how other HN'ers have approached the planning/viability study of getting solar installed at their home. I want to s

by dotBen 6y ago

If it's not off-topic, I'm curious to learn how other HN'ers have approached the planning/viability study of getting solar installed at their home. I want to see a spreadsheet to help me plan out energy use, peak need, whether batteries are worth buying, amount I could sell back, number of panels needed, amortized cost against savings etc. Extra points for thinking tax benefits and estimating the depreciating cost of panels vs the increasing cost of electricity and where the intersection is (or was, if passed) to pull the trigger on a capital expenditure, etc My initial foray so far has been vendors see this as a qualitative decision rather than a quantitative decision with little support for this approach. For me it's purely a numbers game. Anyway, I'm sure HN'ers are most likely to be thinking similarly - any pointers or resources to help others? The more I think about it, there's probably a startup here...

xwdv 6y ago
From a quantitative perspective it will likely not be worth it.
joshocar 6y ago
A lot of it comes down to the individual State incentives. My co-worker put in a system on his house and at the time the incentive was that you sold all of your power back to the grid at a fixed rate that was higher than what you would normally pay. He gets a check every month for what he sells to the grid. That rate was locked in for 15 years. It meant that he knew that with average usage he would break even in 6-7 years. Downside is that he is taxed on that income. The other side of it is how big your roof and if it faces south without much occlusion. If your roof doesn't face south or is occluded but trees then it makes it much less attractive.
fghorow 6y ago
AFAICT, things are already "financialized" so that things like community solar appear to be "cheaper" to a customer. (Nothing wrong with that, BTW.) I just got my first utility bill with a community solar monetary credit. I wrote the following piece attempting to figure out the energy flows -- as opposed to the dollar flows. -- cut here -- OK. I have a(n energetic) head scratcher with my latest unbundled electricity bill. It's the first one that includes a community solar subscription. For geeks, this is almost a thermodynamics question about systems and surroundings -- but it's camouflaged as a monetary problem in disguise... 1) My utility company (NYSEG) charges me a per kW-hr fee for "transmitting" electricity and servicing their infrastructure. Fine and dandy. I understand that perfectly. 2) There is a separate per kW-hr fee from a different supplier of electricity with whom I have contracted for power. Again, I understand that perfectly. 3) Now the head scratcher. My "community solar" account just started contributing PV energy to the grid from a recently completed solar farm. My fraction of energy generated from their deployment just showed up as a monetary credit to my total bill. They will invoice me of 90% of that credit, and claim that I "saved 10%" on my electricity cost. There is no statement whatsoever of the amount of electrical energy corresponding to that monetary credit. That's the monetary accounting. Now let's examine the energy accounting. Let's call the amount of energy from 2) above "N" for my "normal" source of electrical power. Similarly, let us call the (unknown) amount of energy from 3) above "S" for my solar PV power. (In principle, once I knew the fee per kW-hr I was being charged for the PV energy, I could convert the monetary charge to the numerical value for S. Let's leave that as an exercise to be dealt with later after I receive an invoice from the community solar provider -- which hopefully will tell me their cost-per-unit of energy.) Clearly, I paid my normal provider for producing N units of energy. However, I actually consumed (N-S) units of their energy, because I (notionally) also received S units of solar energy. I understand perfectly that electricity on a transmission line is fungible, hence I don't actually receive "electrons" (so to speak) from either of them. The community solar project will also get paid (at their claimed "discounted" rate) for producing S units of energy and delivering it to me via NYSEG. My conundrum: It seems to me that (N+S) units of electrical energy have been produced and paid for, but only N units have been consumed. Hence there are (N+S) - N = S units of electrical energy "somewhere" that are generated but not consumed via this accounting system. Where are those S units? Who owns them? Are they simply going towards heating up the transmission lines, or is somebody benefiting from them? Is it simply the case that NYSEG no longer has to pay for the energy consumed in heating up their transmission lines, but are still charging me the same transmission fees??? I don't know the answer to this problem. Discuss.
- staktrace 6y ago
  Maybe I'm misunderstanding your problem statement. But it seems to me that you are using (and paying for) N units from your regular provider. And then your community solar operation is selling S units to the grid and giving you a monetary credit (the result of the sale) instead. Some customer (due to fungibility, it could be you or anyone else) is getting those S units, and paying the community solar project (indirectly via the grid). And those S units are displacing some other set of S units that were previously generated by fossil fuels, which is (presumably) one of the main goals of this entire exercise. So maybe the answer to your final question is actually "only N units were actually generated, because the fossil fuel plants ran less and produced S fewer units".
  - fghorow 6y ago
    Your take is more sensible than the one I had. The community solar sells S units to somebody else, credits me with the monetary proceeds of the sale (and possibly a share of the associated Renewable Energy Credit proceeds), and charges me 90% of that "accounting fiction". They make money off the difference between the rates they invoice me for and the wholesale rates they actually sell the energy for (via a PPA or something). I still argue that (N+S) units of energy are actually generated. The original generator of N units has no knowledge of the community solar arrangement, and is producing the N units of energy. Unless the community solar is lying about producing S units of energy -- remember, they are never disclosing S, only the associated monetary amount -- they too are selling S units to somebody else at a wholesale rate but getting their cash flow augmented by invoicing me. What a tangled web of cashflows! I think that a PV array on the roof is preferable in situations where a purchaser is able to do that. Less financial engineering, and more physics! YMMV.
zemvpferreira 6y ago
My parents own an off-grid home with ~12 panels (about 3.3kW installed power I think). If it didn't cost over 30K to pull power to the house, it'd be impossible to make the math work for solar. If you don't count the climate externalities I doubt there's one city on Earth where the payback period (before incentives) is less than 10 years. So there's no point in making a spreadsheet at this point, really. You're either buying solar out of need, a desire to "green" your life further, or you're being paid by the government to do it, or buying at a loss.
- mortehu 6y ago
  Why are you treating 10 years as if it's infinity? Lots of houses last longer than 10 years.
  - zemvpferreira 6y ago
    It's just a bad rate of return for an investment. Put the same amount of money in a better investment vehicle (say the stereotypical index fund) and you'll be able to afford buying the same amount of electricity from the grid as the panels would generate and have enough spare to go out to dinner every week.
    - mortehu 6y ago
      You have to compare to investments with similar risk profiles. For example, 10 year treasury bonds only recoup 11.4% of their cost after 10 years. That doesn't mean nobody will accept that return. Not everyone is in a situation where equity index funds make the most sense as a marginal investment.
      - zemvpferreira 6y ago
        Are you also expecting your bonds to be worth literally zero after around 15 to 20 years?
        deleted 6y ago
ip26 6y ago
As you are a numbers guy, I’m sure you can whip up such a spreadsheet in half an hour. It isn’t that hard. NREL has a calculator you can use to estimate annual production, and it’s up to you to figure out what your utility’s rules on net metering are. South is the best aspect, West is next best, but this can be affected by local climate (e.g. frequent afternoon clouds). I think NREL factors climate. If the net metering rules are favorable, skip the battery, forget entirely about your own consumption patterns, and just install as much as you can. The job is almost half fixed costs, so a bigger array pays back much faster.
- jay_kyburz 6y ago
  The hard part in out house is estimating how much power we'll use in the middle of the day, and after we return to the office after covid. The feed-in tariffs here in my state are effectively nothing. I'd like to wait for battery prices to come down so I can go completely off grid. (though this may not happen with cars gobbling up all the batteries for a long time to come) I think as electricity gets cheaper and cheaper, connection costs will get more and more expensive, so being able to disconnect will be a significant saving. Right now we are working to remove all gas appliances here at home so we can disconnect from the Gas entirely.
  - bluGill 6y ago
    If you have lived there for a few years you can look at your bills from years past and be close enough. You will never be more than close enough anyway
marktangotango 6y ago
Basically the numbers come down to a few things. How much energy does your home use in a day (kw/h). How much energy does your installed solar generate; this is genarally kw of solar times average hours of peak sun for your latitude. So if you want to entirely power your home with solar, and you average 1,000 kw/h a month that's about 33 kw/h a day, and your latitude averages 5 hours of direct sun a day, you need about 7 kw of panels to generate that 33 kw/h. Then you need to store the excess power to use throughout the rest of the day. If you're off grid, this means a lot of battery. If you're on grid (grid tie) you can generally put your excess power on the grid, then draw it back out throughout the day and your utility company will "net meter" periodically so you only pay for the difference you used, or they pay you for the excess you didn't use. What utility companies pay for excess residential has evolved and is evolving. A few years ago you could get the full domestic rate (~ $.12 per kw/h) and that is a fantastic rate. Now utilities are pushing for a lower rate to take into account for delivery costs solar does not require [1]. [1] https://insideclimatenews.org/news/11062019/rooftop-solar-net-metering-rates-renewable-energy-homeowners-utility-state-law-changes-map/ https://insideclimatenews.org/news/11062019/rooftop-solar-ne...
ww520 6y ago
I've gone through a similar process. The gist of it is that it takes about 6-7 years to recoup the investment, disregarding compound return and opportunity cost. You can think of it as buying a bond that pays back in 6-7 years and then getting 23-24 more years of free return (discounting the teardown cost eventually). Here's a more detail approach. 1. Compute your electricity usage. Look at 12 months of electric bill. E.g 800kWh per month (9600kWh per year). 2. Get quotes from the solar vendors. They usually quote in term of 5kW system, 6kW system, or 10kW system. The kW number is the instantaneous electricity the panels generated. Multiply that by hours to get the kWh number. The number of panels doesn't matter because some panels generate more some less. You want the total system output number in kW. 3. Figure out how many hours of useful sunlight a day your house can have. Google Sunroof and other calculators on the web would do it for you. E.g. 5 hours of useful sunlight per day. 4. E.g. with a 6kW system, 6kW x 5 hours/day = 30kWh per day = 900kWh per month = 10800kWh per year. Remember you need 9600kWh/year from 1? This just about covers it, accounting for some loss of efficiency. 5. A 6kW system costs about $14,460–$19,260 in CA. See [1]. Let's say it costs $18,000. Let's say your electric bill is $250/month or $3000/year. $18,000/$3,000 = 6 years to recoup (very simplistic view). Battery is an additional cost, $6000~$10000. [1] https://news.energysage.com/6kw-solar-system-compare-prices-installers/ https://news.energysage.com/6kw-solar-system-compare-prices-...
- jay_kyburz 6y ago
  Buy simply comparing your bill with how much you generate is not very useful. You have to know when you used that power, and whether or not the sun was shining at that time of the day. You need to know if your local utility will even pay you for the power you generate that you don't use. If you are out of the house for most of the day, you probably don't use much power then. When you get home at night you turn all your devices on and start cooking and cleaning.
  - ww520 6y ago
    Net metering or battery can timeshift the energy usage.
- bluepoint 6y ago
alkonaut 6y ago
With subsidies I could make a profit in 15 years. But I don't want to make a plan that stretches that long, and that's a best case scenario (assuming no damages, expensive failures). So while I can get 100% renewable from the grid, I'll stick with that, at least until an installation can pay for itself in just a few years.
- ClumsyPilot 6y ago
  I this that with this approach you will forever be buying electricity from those who are happy with a 10-15 year financial plan. There is nothing wrong with it, but i dont see how your expectation of a 98%?safe investment paying for itself in 5 years is justified.
  - alkonaut 6y ago
    What I think will happen is that these installations will become common so they are an investment that adds a near equivalent value to the home. Once that happens, it’s also safe to buy it regardless of whether it takes 15 years or more so I sell the house. The buyer will pay more for my house. That’s not the case yet. Solar doesn’t add a value equivalent to the investment to the house, so unless I plan on staying until it’s paid off - it’s not worth it. But I think this will change.
    - ClumsyPilot 6y ago
      Thats a good point, but what do you recon is the 'added value' currently?

by dotBen 6y ago

xwdv 6y ago

From a quantitative perspective it will likely not be worth it.

joshocar 6y ago

A lot of it comes down to the individual State incentives. My co-worker put in a system on his house and at the time the incentive was that you sold all of your power back to the grid at a fixed rate that was higher than what you would normally pay. He gets a check every month for what he sells to the grid. That rate was locked in for 15 years. It meant that he knew that with average usage he would break even in 6-7 years. Downside is that he is taxed on that income. The other side of it is how big your roof and if it faces south without much occlusion. If your roof doesn't face south or is occluded but trees then it makes it much less attractive.

fghorow 6y ago

AFAICT, things are already "financialized" so that things like community solar appear to be "cheaper" to a customer. (Nothing wrong with that, BTW.) I just got my first utility bill with a community solar monetary credit. I wrote the following piece attempting to figure out the energy flows -- as opposed to the dollar flows. -- cut here -- OK. I have a(n energetic) head scratcher with my latest unbundled electricity bill. It's the first one that includes a community solar subscription. For geeks, this is almost a thermodynamics question about systems and surroundings -- but it's camouflaged as a monetary problem in disguise... 1) My utility company (NYSEG) charges me a per kW-hr fee for "transmitting" electricity and servicing their infrastructure. Fine and dandy. I understand that perfectly. 2) There is a separate per kW-hr fee from a different supplier of electricity with whom I have contracted for power. Again, I understand that perfectly. 3) Now the head scratcher. My "community solar" account just started contributing PV energy to the grid from a recently completed solar farm. My fraction of energy generated from their deployment just showed up as a monetary credit to my total bill. They will invoice me of 90% of that credit, and claim that I "saved 10%" on my electricity cost. There is no statement whatsoever of the amount of electrical energy corresponding to that monetary credit. That's the monetary accounting. Now let's examine the energy accounting. Let's call the amount of energy from 2) above "N" for my "normal" source of electrical power. Similarly, let us call the (unknown) amount of energy from 3) above "S" for my solar PV power. (In principle, once I knew the fee per kW-hr I was being charged for the PV energy, I could convert the monetary charge to the numerical value for S. Let's leave that as an exercise to be dealt with later after I receive an invoice from the community solar provider -- which hopefully will tell me their cost-per-unit of energy.) Clearly, I paid my normal provider for producing N units of energy. However, I actually consumed (N-S) units of their energy, because I (notionally) also received S units of solar energy. I understand perfectly that electricity on a transmission line is fungible, hence I don't actually receive "electrons" (so to speak) from either of them. The community solar project will also get paid (at their claimed "discounted" rate) for producing S units of energy and delivering it to me via NYSEG. My conundrum: It seems to me that (N+S) units of electrical energy have been produced and paid for, but only N units have been consumed. Hence there are (N+S) - N = S units of electrical energy "somewhere" that are generated but not consumed via this accounting system. Where are those S units? Who owns them? Are they simply going towards heating up the transmission lines, or is somebody benefiting from them? Is it simply the case that NYSEG no longer has to pay for the energy consumed in heating up their transmission lines, but are still charging me the same transmission fees??? I don't know the answer to this problem. Discuss.

staktrace 6y ago
Maybe I'm misunderstanding your problem statement. But it seems to me that you are using (and paying for) N units from your regular provider. And then your community solar operation is selling S units to the grid and giving you a monetary credit (the result of the sale) instead. Some customer (due to fungibility, it could be you or anyone else) is getting those S units, and paying the community solar project (indirectly via the grid). And those S units are displacing some other set of S units that were previously generated by fossil fuels, which is (presumably) one of the main goals of this entire exercise. So maybe the answer to your final question is actually "only N units were actually generated, because the fossil fuel plants ran less and produced S fewer units".
- fghorow 6y ago
  Your take is more sensible than the one I had. The community solar sells S units to somebody else, credits me with the monetary proceeds of the sale (and possibly a share of the associated Renewable Energy Credit proceeds), and charges me 90% of that "accounting fiction". They make money off the difference between the rates they invoice me for and the wholesale rates they actually sell the energy for (via a PPA or something). I still argue that (N+S) units of energy are actually generated. The original generator of N units has no knowledge of the community solar arrangement, and is producing the N units of energy. Unless the community solar is lying about producing S units of energy -- remember, they are never disclosing S, only the associated monetary amount -- they too are selling S units to somebody else at a wholesale rate but getting their cash flow augmented by invoicing me. What a tangled web of cashflows! I think that a PV array on the roof is preferable in situations where a purchaser is able to do that. Less financial engineering, and more physics! YMMV.

zemvpferreira 6y ago

My parents own an off-grid home with ~12 panels (about 3.3kW installed power I think). If it didn't cost over 30K to pull power to the house, it'd be impossible to make the math work for solar. If you don't count the climate externalities I doubt there's one city on Earth where the payback period (before incentives) is less than 10 years. So there's no point in making a spreadsheet at this point, really. You're either buying solar out of need, a desire to "green" your life further, or you're being paid by the government to do it, or buying at a loss.

mortehu 6y ago
Why are you treating 10 years as if it's infinity? Lots of houses last longer than 10 years.
- zemvpferreira 6y ago
  It's just a bad rate of return for an investment. Put the same amount of money in a better investment vehicle (say the stereotypical index fund) and you'll be able to afford buying the same amount of electricity from the grid as the panels would generate and have enough spare to go out to dinner every week.
  - mortehu 6y ago
    You have to compare to investments with similar risk profiles. For example, 10 year treasury bonds only recoup 11.4% of their cost after 10 years. That doesn't mean nobody will accept that return. Not everyone is in a situation where equity index funds make the most sense as a marginal investment.
    - zemvpferreira 6y ago
      Are you also expecting your bonds to be worth literally zero after around 15 to 20 years?
      - deleted 6y ago

ip26 6y ago

As you are a numbers guy, I’m sure you can whip up such a spreadsheet in half an hour. It isn’t that hard. NREL has a calculator you can use to estimate annual production, and it’s up to you to figure out what your utility’s rules on net metering are. South is the best aspect, West is next best, but this can be affected by local climate (e.g. frequent afternoon clouds). I think NREL factors climate. If the net metering rules are favorable, skip the battery, forget entirely about your own consumption patterns, and just install as much as you can. The job is almost half fixed costs, so a bigger array pays back much faster.

jay_kyburz 6y ago
The hard part in out house is estimating how much power we'll use in the middle of the day, and after we return to the office after covid. The feed-in tariffs here in my state are effectively nothing. I'd like to wait for battery prices to come down so I can go completely off grid. (though this may not happen with cars gobbling up all the batteries for a long time to come) I think as electricity gets cheaper and cheaper, connection costs will get more and more expensive, so being able to disconnect will be a significant saving. Right now we are working to remove all gas appliances here at home so we can disconnect from the Gas entirely.
- bluGill 6y ago
  If you have lived there for a few years you can look at your bills from years past and be close enough. You will never be more than close enough anyway

marktangotango 6y ago

Basically the numbers come down to a few things. How much energy does your home use in a day (kw/h). How much energy does your installed solar generate; this is genarally kw of solar times average hours of peak sun for your latitude. So if you want to entirely power your home with solar, and you average 1,000 kw/h a month that's about 33 kw/h a day, and your latitude averages 5 hours of direct sun a day, you need about 7 kw of panels to generate that 33 kw/h. Then you need to store the excess power to use throughout the rest of the day. If you're off grid, this means a lot of battery. If you're on grid (grid tie) you can generally put your excess power on the grid, then draw it back out throughout the day and your utility company will "net meter" periodically so you only pay for the difference you used, or they pay you for the excess you didn't use. What utility companies pay for excess residential has evolved and is evolving. A few years ago you could get the full domestic rate (~ $.12 per kw/h) and that is a fantastic rate. Now utilities are pushing for a lower rate to take into account for delivery costs solar does not require [1]. [1] https://insideclimatenews.org/news/11062019/rooftop-solar-net-metering-rates-renewable-energy-homeowners-utility-state-law-changes-map/ https://insideclimatenews.org/news/11062019/rooftop-solar-ne...

ww520 6y ago

I've gone through a similar process. The gist of it is that it takes about 6-7 years to recoup the investment, disregarding compound return and opportunity cost. You can think of it as buying a bond that pays back in 6-7 years and then getting 23-24 more years of free return (discounting the teardown cost eventually). Here's a more detail approach. 1. Compute your electricity usage. Look at 12 months of electric bill. E.g 800kWh per month (9600kWh per year). 2. Get quotes from the solar vendors. They usually quote in term of 5kW system, 6kW system, or 10kW system. The kW number is the instantaneous electricity the panels generated. Multiply that by hours to get the kWh number. The number of panels doesn't matter because some panels generate more some less. You want the total system output number in kW. 3. Figure out how many hours of useful sunlight a day your house can have. Google Sunroof and other calculators on the web would do it for you. E.g. 5 hours of useful sunlight per day. 4. E.g. with a 6kW system, 6kW x 5 hours/day = 30kWh per day = 900kWh per month = 10800kWh per year. Remember you need 9600kWh/year from 1? This just about covers it, accounting for some loss of efficiency. 5. A 6kW system costs about $14,460–$19,260 in CA. See [1]. Let's say it costs $18,000. Let's say your electric bill is $250/month or $3000/year. $18,000/$3,000 = 6 years to recoup (very simplistic view). Battery is an additional cost, $6000~$10000. [1] https://news.energysage.com/6kw-solar-system-compare-prices-installers/ https://news.energysage.com/6kw-solar-system-compare-prices-...

jay_kyburz 6y ago
Buy simply comparing your bill with how much you generate is not very useful. You have to know when you used that power, and whether or not the sun was shining at that time of the day. You need to know if your local utility will even pay you for the power you generate that you don't use. If you are out of the house for most of the day, you probably don't use much power then. When you get home at night you turn all your devices on and start cooking and cleaning.
- ww520 6y ago
  Net metering or battery can timeshift the energy usage.
bluepoint 6y ago

alkonaut 6y ago

With subsidies I could make a profit in 15 years. But I don't want to make a plan that stretches that long, and that's a best case scenario (assuming no damages, expensive failures). So while I can get 100% renewable from the grid, I'll stick with that, at least until an installation can pay for itself in just a few years.

ClumsyPilot 6y ago
I this that with this approach you will forever be buying electricity from those who are happy with a 10-15 year financial plan. There is nothing wrong with it, but i dont see how your expectation of a 98%?safe investment paying for itself in 5 years is justified.
- alkonaut 6y ago
  What I think will happen is that these installations will become common so they are an investment that adds a near equivalent value to the home. Once that happens, it’s also safe to buy it regardless of whether it takes 15 years or more so I sell the house. The buyer will pay more for my house. That’s not the case yet. Solar doesn’t add a value equivalent to the investment to the house, so unless I plan on staying until it’s paid off - it’s not worth it. But I think this will change.
  - ClumsyPilot 6y ago
    Thats a good point, but what do you recon is the 'added value' currently?