4 ms·
The point is the homeowner saves money from day one, not even taking into account current rebates. The utility bill savings more then compensate for the extra
by marze 15y ago
The point is the homeowner saves money from day one, not even taking into account current rebates.
The utility bill savings more then compensate for the extra $60 in the monthly loan payment.
Most solar panels come with a 25 year warranty.
- anamax 15y ago> Most solar panels come with a 25 year warranty. How much of "rated power" does that warranty cover? Suppose that the power output drops 20% by year 29, what will the warranty pay? How about if the output drops 15% by year 20? It's unlikely that the warranty will pay full replacement value. Most warrantys cover pay value, which is likely to be fairly low for 20-30 year-old PV panels. And no, they won't fix if that costs more than the "value".
- justincormack 15y agoIt is unlikely the warranties would pay for that, the manufacturers couldnt afford to basically replace panels for free.
- lsc 15y agodepends entirely on how much they expect costs to drop. For instance, I would be very happy to provide you a 10 year warranty on 2tb of storage. Sure, it's going to fail and I'm going to have to replace it, but if I can get you to pay a price that would seem fair to replace the thing once or twice at today's prices, most likely when your device does fail, I'll be able to replace it with much cheaper parts. (Of course, this ignores the format differences; if sata goes out of style, it could actually cost me more to source a sata part. Of course, judging from how long PATA lasted in a largely backwards compatible format, I /probably/ wouldn't have to worry about that too much in the next 10 years, but it'd be a source of worry. I'd be much more comfortable promising you 2tb of storage of some type with some minimum performance level, and not specifying interface, or selling you network-accessible storage for the next ten years.)
- richardw 15y agoThe money saved would reduce substantially if you consider the economic cost of investing the $12500 over 30 years. My guess is the best approach would be to invest the money until the price of solar stops dropping so quickly. Edit: I'd love to put solar in my house. We have a rebate in my country as well, but I'm sure I can use the money better in the short term, and likely end up with a more efficient setup for less money if I wait. Until then, geyser blanket, install LED's in the most-used light fittings, install a small geyser under the kitchen sink so it doesn't draw the full pipe's water just to wash dishes.
- roel_v 15y agoWhat kind of rebate do you have? According to my calculations one needs very little subsidies to make installing solar panels attractive financially, even when discounting for opportunity cost (of course if you take a yield of 7% on your money then it's different, but who has made 7% on small amounts (10-20k) over the last few years and is willing to use that yield as an assumption for the next 2 decades?)
- richardw 15y agoIIRC approximately 15%, which is a lower than it was a year ago. Prices range up to 3x the lowest price, with a suggested 2x to be sure of a good quality setup and 10 year guarantee (which seems short). However, if you buy a system now and it's 10% cheaper in a year and you would have made say 3.5% on the investment [1] you'll be out 13.5% minus your electricity savings, and that's in 1 year. What about 4-5 years? Won't it be much cheaper then? [1] I've made about 37% this year as a fairly active trader, so every cent means a margin of safety. No solar for me because I need the cash!
- roel_v 15y agoI've made the numbers match (i.e., same cost for PV installation or using power from grid, assuming defensive increases in cost of electricity for the next 15 years - only going up as much as inflation) with a slightly higher subsidy (1000€ subsidy on an installation cost of a little under 12k€, which would only buy an installation that doesn't produce 100% of an average consumption). But when the costs converge this closely then small assumptions can tip the scale one way or the other - one % extra opportunity cost, a converter that needs replacing 5 years earlier, 10% less sunshine in one of the earlier years of the installation's lifespan, ... That said though, when you extrapolate costs of grid electricity of the last 10 years, and base your energy price scenario on that (instead of keeping up with inflation like I did), PV comes out ahead even when taking worst-case scenarios of all other variables. Many people argue that because prices are falling so fast, it doesn't make sense to buy now. But IMO it's much more nuanced than that: - As prices fall, so will subsidies. Already most subsidies are being phased out or decreased. - There is a floor on the prices - in the short term that floor is just below grid parity, in the medium and longer term it will be production limits. I don't think it's rational to assume that prices will keep falling so fast. Many words have been shed on this - I guess only time will tell. That said, like I posted elsewhere here, my parents are getting an ROI of 11% on their panels. That's mighty good for money that can be borrowed under mortgage, and 'green' == cheap mortgages for that matter. (that yield is only because of subsidies though, so it's not really a fair general case comparison).