Slashdot Mirror

"A 2000 sq ft house in suburban Arizona can power itself from PV cells on the roof"

Maybe, maybe not. Probably not without centralized generation for the demand surges.

Note the thread I linked to goes off on tangents, offering examples from Long Beach and Florida. Add 20 degrees ambient, forego A/C in all ancillary areas, and size your A/C system conservatively (properly according to my A/C guy friend), and you mostly can. Huge investment in systems. What is the ROC/NPV of this? Not to mention the conversion losses, but the Sun is free-ish.

Deep cycle batteries are designed to be discharged down as much as 80% time after time, and have much thicker plates. The major difference between a true deep cycle battery and others is that the plates are SOLID Lead plates - not sponge. This gives less surface area, thus less "instant" power like starting batteries need. Although these can be cycled down to 20% charge, the best lifespan vs cost method is to keep the average cycle at about 50% discharge. Unfortunately, it is often impossible to tell what you are really buying in some of the discount stores or places that specialize in automotive batteries. The golf car battery is quite popular for small systems and RV's.

http://www.solar-electric.com/...

A small lead acid bank capable of driving a house for an evening isn't that expensive.

Please define "not that expensive".

Eight of these for example would run you just about $3000 (I chose 8 to get you to a 48V system, the common configuration for off-grid systems these days). I include shipping but not the wiring and interconnects and such to make them work for you:

http://www.wholesalesolar.com/...

A lead-acid system shouldn't be discharged more than around 20% in order to avoid longer term damage; you can discharge them more deeply but you'll shorten their life significantly:

http://www.solar-electric.com/... of Batteries

Scroll down to the "Cycles vs. Lifespan" section and click on that chart. It's amazing.

So if you are likely to use half of your storage (~225AH for those selected) then a single set of batts would be discharged 50% in normal use. At that rate the batts will last you around 1000 charge/discharge cycles, so a bit less than 3 years if it's every day. If you want them to only be discharged 75% (moves you up the lifetime curve to 2000 charge/discharge cycles or around 5.8 years) it'll cost you double the price above, or around $6K. And quite honestly 225AH is a very shallow system....we're not talking about much being on overnight here. You definitely couldn't consider running a microwave--you'd tap that puppy right out.

I'm leaving out the equipment costs and such; charge controllers to dump power into the batteries and an inverter to take power out will add around $5K more.

I love the idea of as many folks as possible having back up and/or being off-grid entirely (I am) but it ain't cheap. I did it because I had no other choice here; the house is 5 miles away from the nearest power line.

Ferret

I have put in driveway snowmelt systems and a typically driveway needs at a minimum ~100 kbtu/hr boiler to keep the driveway clear. Scaling that up to a road way and it would be astronomical.

That was my thought as well. Phase change is a bitch, so I anticipated this was a marketing gimmick. I decided to run some quick calculations to determine how much snow could be melted by a 1 m^2 solar heating roadway plate thing.

Solar Roadways is in Idaho, so I decided to use their location for stats. I decided to use an average insolation value of 2 kWh/day in December in Idaho. I disregarded the fact that these plates won't be tilted to compensate for latitude, which will give the roadway an artificially improved performance stat. I used an enthalpy of fusion for water as 334 kJ/kg. I used a 50 kg/m^3 value for the density of freshly-fallen snow. Finally, I decided to let the road panel have a 15% PV efficiency as well as a 100% solar panel coverage (neither of which is likely to be realistic for a road tile thing, but again this is in favor of the roadway panel).

So, how much snow can this melt per day? Call it 6.5 cm. In practice, I'm guessing the answer is closer to "0", because the instant the panel is covered by snow it will cease generating energy. Also, snowstorms are not known to occur during bright, bright, sunshiny days. It seems Solar Roadways expects their panels to be hooked to the grid and pull power to melt snow.

Therefore, this exercise devolves to "why haven't we installed electric radiant heat in our existing roadways to melt snow?"

Well, if we have a four lane standard US highway (12 ft lanes) and we need to melt that same 6.5 cm of freshly fallen snow, it would require 4.4 MWh (yes, megawatt-hours). In Idaho, it looks like an average wholesale rate for 1 MWh of electricity is approximately $150. So... call it $600 per km to melt a few cm of snow... once? And this is for light, fluffy, happy snow, not the slushy sleety shit that has the density of neutronium and gives grandpa a heart attack when he tries to shovel it.

Unless I dropped a few orders of magnitude here (please let me know if I did), it seems the answer to this is "just use the fucking salt instead, like we have been doing." In conclusion, perhaps the LED roadway is useful, but the snow melting bit really seems to be a gimmick.

Since the switch would need a battery and charger of some sort to operate it and the radio link, just get a pump that runs on solar.
They make anything from .25GPM up to like 50,000GPD.

solar pumps

You're the second person to talk about Deep Cycle batteries but seemingly only read the name. If you follow even the wikipedia link you will find it saying they can be discharged up to 80% for several cycles, but the reality is you don't want to routinely discharge them that deep. They will deliver far more watt hours discharged to only 50% or less.

"Deep cycle batteries are designed to be discharged down as much as 80% time after time, and have much thicker plates."

As for the payback period (not sure why a link to wikipedia here)

The link to the wiki article wasn't about the payback period. If you had read it you would have seen it was about ROI, Return on Invest, which is what the text for the link was. Not everyone knows what ROI is and for those who don't I included a link to a wiki article explaining what it is.

even grid-tie solar panel systems never pay for themselves

I said nothing about grid tied systems in the post you replied to, but I did specifically mention about "off the grid" systems. Most of those who do build off the grid install a solar, wind, or hybrid system specifically because it does cost less. To have the power company install cabling just a few thousand feet can cost more than $10,000. You either pay that plus a months power bill or you use the money to buy and build a solar system then you're free of having to pay until it or parts of it are replaced. All it requires is a little maintenance.

"Since the house is about 1,200 ft from the Arizona Public Service grid, it qualified for utility service, although connection would have cost about $5,700."

"Sunny and her former husband bought the property 18 years ago and spent a few thousand dollars on a solar power system. Connecting to the power grid would have cost $80,000, but Sunny, 53, had no interest anyway."

The costs are easy to check. Panels in the 100-200W class run about $5/W or a little less. That's consumer price, small quantity. Pick your favorite solar panel store; I looked at several panels from Northern Arizona Wind and Sun. That metric is about the most conservative I can think of, and it still justifies PV panels as energy positive. I haven't looked at the study you mention, but many such studies ignore things like the higher labor costs involved in PV compared to grid power -- just because that money doesn't get spent on energy that goes into the PV panels directly doesn't mean it has no carbon impact. I don't know how much is needed to correct for that, but I do know my method is about as conservative as possible.

$1000 for 160W is high. 170W, $845 when buying 12+
Even the MSRP is only $993. $5/W is usually a close enough price for these panels, less if you buy in bulk.
I agree with most of the rest of your post, however they probably are not adding additional batteries (most likely whatever they are running already has a UPS/generator system in place).
possibly $300,000 each for a 500kW inverter if you have 16 of them that's only $4.8 million
So lets say $40mil for panels, $10mil for inverters, cables, installation, etc, thats $50 mil total, or .038% of their market cap, or 3% of YouTube

This is so far off topic, but I can't help but respond (I hate HOAs more than I hate SUVs, or nuclear power for that matter)

As posted here, it's essentially a contract attached to the property, and they're generally enforceable to the same extent that contracts are enforceable, with the added caveat that states can regulate what can and cannot be enforced. Most of the discriminatory restrictions like barring blacks and such have been struck down by courts, and thought they can still be found in CCRs, they're unenforceable. I remember something with somebody wanting to put up a flag pole post Sept. 11th in AZ, and the HOA raising a fuss, and so the state legislature passed a law specifically stating that they couldn't ban flying U.S. flags, as well as forbidding HOAs from banning campaign signs:

In Arizona, lawmakers responded to the practice of boards meeting outside the state -- an attempt to keep members from voting -- by passing a bill requiring those gatherings to be held in Arizona.
Update: In 2004, the Arizona State Legislature amended existing legislation to protect homeowners' and condominium residents' right to fly flags. The revised statues (33-1261 and 33-1808) allow for residents to fly flags that are consistent in size with dimensions noted in the federal flag code (P.L. 94-344; 90 Stat. 810; 4 U.S. Code sections 4-10). Arizona homeowners' associations may not prohibit the installation of a flagpole, but may dictate its height and placement.

The revised statute 33-1808, which concerns homeowners in planned communities, also secured the right to post political signs. Signs may be displayed unless regulated by the association. All sign prohibitions must be consistent with and not exceed restrictions in local sign ordinances. Political signs may never be prohibited 45 days before or 7 days after an election.

Full article here

Incidentally, AZ law also forbids HOAs from banning solar panels on homes (more info here and here)

The problem with HOAs is that they are private little governments that are unencumbered by many of the constitutional restrictions that the federal and state governments have to follow. In response to the person who said "well, don't move into a neighborhood with a HOA," I'd like to point out that it's getting increasingly difficult to do that. Almost every new house built in the last 10-20 years is covered by a HOA because people are silly enough to believe their hype that they'll keep neighborhoods clean and safe and protect property values (property values that would be better protected if they actually built quality homes and better neighborhoods). In fast-growing cities like Phoenix, that's a huge percentage of the housing stock (I've read as high as 50% of all houses in metro Phoenix).

OK, I can comment a little bit on this. First of all, there is a hosting company that claims to run entirely on renewable energy (solar, wind, they don't get to specific on their site), SolarHost.com. I've never used them, only read about them.

As for running on solar power (or wind), what you want to consider to start with is how much power the box draws, and how long it will run. The best thing you can do is conserve what you can; it will save you on overall system cost. If it's a workstation, pick one just as big as you need, and don't go overboard. (I'm typing this on my dual AMD MP box, which yes is overkill. But in case anyone's interested, with both CPUs at 100 % doing some Maya rendering among other things, it draws about 250 watts.) A server is the same way, but you'll usually want to leave it on all the time. You can take many conservation measures, like making it a headless box and minimizing the hardware involved. My Gentoo Web/Mail/MySQL server at home is a 1 GHz Via board with 256 MB Ram, and the 12 volt car power supply. All I have connected to the board is a hard drive; I disconnected the CD-ROM when I was done installing. At 12 volts, it draws roughly 30-36 watts (~3 amps).

Going the 12 volt route is nice, because then you don't need an inverter to power the box. Assuming it draws 35 watts 24 hours/day, that comes out to 840 watt-hours a day (you get watt-hours by multiplying watts by hours). Obviously, we're going to want to use a battery so our server can run at night. So, we multiply by 1.2 to compensate for the power lost here, and get 1008 watt-hours (1.008 KWH). Now, assuming we want to do this entirely with solar, we need to figure out how much sun is available in the area. Go here to figure out how many hours of sun you get a day. Where I am, in New York State, we get an average of 2.5-3 sun hours a day throughout the year (keep in mind, this is "full sun" so the panels will be putting out power at different light levels, though well stick with this figure of the sun hours). Let's use 2.5 sun hours to be safe. To produce 1008 watt-hours, we divide 1008 by 2.5. The result is that I would need 403.2 watts of solar panels. Four of these will be about right, putting out 115 watts each or about 460 watts, leaving a bit of a buffer. The price of the panels alone would be $2,060. You'd need batteries next, with enough capacity for, let's say, 3 days without sun. That means they have to be able to hold 3.024 KWH at 12 volts, or 252 amp-hours (3,024 watt-hours divided by 12 volts). If we drain the battery (assuming a deep-cycle led-acid battery) down all the way, we'll damage it. The minimum should be 20 percent. To make sure they last long, however, we should go with something more along the lines of, say, 35 %. That means we should plan for a battery bank with a capacity of about 390 amp-hours. A pair of the L-16 HCs found here should work. Then you'd need a charge controller of about 50 or 60 amps. Not including wiring, labor, fusing, shipping, or the server itself, the cost comes to about $3,116. This is to run a 35 watt load 24 hours a day.

Believe it or not, small-scale renewable energy is in fact quite popular. A lot can be done with the power generated by that system; you could probably run a small apartment with it (assuming no heat with electricity, and that said apartment does not contain the server :-)). Most people don't run a server all the time, so something like this works out. Running an average-sized house, after some slight conservation (not using electricity for heat, using compact floursecent lights and LEDs, and just remembering to turn stuff off when not using it), the total is usually about $20,000 to $30,

A decent solar panel will also set you back anywhere between $50k and $500k. And even with a $500k panel, one gets only 3-4kW out of it under best circumstances, barely enough to sustain an effort in the 4-5HP area.

I don't know where you got those numbers but solar panels are a lot cheaper than that. Doing a real quick google I found this:

Isofoton 150W 24 volt 25 amp solar panels for $675.00 each or 4 for $2,600.00. Using $2,600 for four panels 20 panels wired in series will generate 3kw for $13,000.

That's nowhere near $50k you said for one panel. Having said that, it's cheaper and easier to conserve. One way to conserve is by changing light bulbs to CFL, compact floresent lights, they use 1/4 of what "regular" incandecent lights do and provide just as much lighting. That's what I use, 12 and 15 watt CFLs that illuminate as much as 60 and 75 watt incandecent light bulbs. LED light bulbs use even less energy. There are LED light bulbs that use less than 10% of the energy an incandecent and still provide as much light.

To address some of Klevin's confusion, since I've been following solar panel advancements:

Thin, flexible cells have been around for a while. One reason they haven't caught on heavily is because they're nowhere near as 'powerful' (efficient at conversion) as hard panels. Did a quick search (don't take this data *too* seriously, but it represents what's normal); compare panels from these two pages:
Flexible
Solid

Specifically, compare "Unisolar 32 watt flexible solar panel" from the first link to "Shell ST40 thin film CIS 40 watt solar panel" on the second. The flexible panel is 940 sq. inches and 32 watts, while the solid panel 663 sq. inches and 40 watts. Big difference in watt per area.

I ended up choosing a big solid one to fit in the rear dash of my car; flexible would have been easier to deal with, but it won't fold, and produces less power. (I use the panel in my car to power my laptop/cell phone combo while camping and stuff, it's very cool and gets a lot of questions from random interested people!)

Here's another chart to compare the two: Product Page
Tried to find an efficiency rating chart comparing the two types, but no luck. The numbers are out there somewhere...

To address some of Klevin's confusion, since I've been following solar panel advancements:

Thin, flexible cells have been around for a while. One reason they haven't caught on heavily is because they're nowhere near as 'powerful' (efficient at conversion) as hard panels. Did a quick search (don't take this data *too* seriously, but it represents what's normal); compare panels from these two pages:
Flexible
Solid

Specifically, compare "Unisolar 32 watt flexible solar panel" from the first link to "Shell ST40 thin film CIS 40 watt solar panel" on the second. The flexible panel is 940 sq. inches and 32 watts, while the solid panel 663 sq. inches and 40 watts. Big difference in watt per area.

I ended up choosing a big solid one to fit in the rear dash of my car; flexible would have been easier to deal with, but it won't fold, and produces less power. (I use the panel in my car to power my laptop/cell phone combo while camping and stuff, it's very cool and gets a lot of questions from random interested people!)

Here's another chart to compare the two: Product Page
Tried to find an efficiency rating chart comparing the two types, but no luck. The numbers are out there somewhere...

Many green energy solutions can be implemented in a decentralized manner, instead of in huge projects like you mention. Installing them in this manner could make use of space that could not otherwise be used for energy generation.

Covering your house's shingles with solar panels would be expensive, but this could provide for much of the electrical needs of your household. For about $14,000, you can buy 24 165w Sharp 1575mm x 826mm solar panels, and save about $500 a year on electricity.

A 20m tower with a 7m diameter wind turbine could be installed in even a very small inner-city house lot. If you live in a reasonably windy climate, this could generate all the electricity you need for about $25,000.

I know these are expensive solutions, but certainly not impossible. The prices will come down.

it's nice that you get a warm fuzzy inside from all these ideas, but there are a lot of hidden costs.

first off, a wind genny that's going to actually give back to the grid seems to cost more than 50k. while nice for the environment, it's not very practical in any sense. that's 1/2 the house you describe above.

regarding the house you describe. sure the land/house you get will be a few grand cheeper, but everything is about location. a house like that is definately out in the sticks. you gain peace and quiet, and it costs you the convenience of living near a city. so, the option you describe is only valid if you're planning to live in the sticks. that's a small % of the folks there.

regarding the "other" items in our lives: tv, boat, gaming system, etc. these are entertainment items. these are purchased with entertainment dollars. typically what's left after expenses and savings becomes available for entertainment/recreational purposes.

electricity, heat, cooling, gasoline are all expenses. thus they're going to come from expense dollars. in order for the masses to flock to any other technology, it must be more cost effective in the long run. i can't even get a 20 year payback on the investment of a self sufficient solar system for my house. that's not going to cause people to flock there in masses. maybe it's a chicken/egg problem. people aren't buying the stuff, so manufacturing isn't huge, so prices are high. maybe raw materials to make these solar panels are high, but i doubt that. i've read you can kinda build one from common materials. haven't tried yet.

i also burn wood all winter long. gas bills are extremely low (have a innefficient water heater, but again, replacing it would either have to be cost effective, or it'll have to die). the wood is essentially free, even in the city. that's only because wood buring use is low and tree cutting business is healthy. the tree companies cut more trees than they can sell as firewood. so, i have a splitter and a hauler to go and grab what ever they're willing to give away. now, if i figure in the HOURS i spend on the wood (splitting cutting, hauling stacking, etc, it definately doesn't make financial sense. put my time at a measly 10$ per hour, and the gas heat is going to be cheeper. maybe i'm a slow cutter/splitter. but there's also the time to carry it in the house ever other day. and time feeding the fire, and getting it started after a days work. i love it, but it's a lot of work and it's not cost effective. thus, most people aren't going to use it.

if i decide to build a place out in the hills where land is cheep, i'll definately consider geothermal, and maybe even solar and even wind power. geothermal is the one that seems to have the best payoff. 10-15 years is what i seem to gather, and then it's money in the bank. you have less to worry about adding 15k to the value of your house that way.

Thanks for the reply.

I'm looking into similar gear, minus the solar currently. I'm planning on putting in a Trace DR series 3600 watt inverter/charger/UPS with a 200-400 amp hour bank. Probably wire it up into a subpanel and put some circuits directly on it. We get frequent power outages from storms and such where I live. The rationale is "Why spend $1200 for three 400 watt UPSes with expensive SLA batteries when I can get something like the Trace and put as many flooded batteries as I want on it."

I was amazed how cheap something like the Trace DR is compared to the equivalent computer UPS from APC. IIRC both are modified sine output.

BTW- Did trace change their name to Xantrex? That sounds like some prescription drug. :)