|
Claverjoe posted:Oh, a direct solar water heater instead of a solar electric-to-heat would be a lot more efficient for you. Honestly, solar water heaters are the better deal if you are starting out with just needing things like hot water and radiator heating. The problem with direct solar water heat is that I have a system in place, it's central and heat's 3 buildings, including hot water in all 3. The system has been in place for over a decade and is actually quite efficient on it's own. I just want to reduce the amount I burn.
|
#
?
Feb 25, 2013 23:07
|
|
|
|
| # ? Apr 25, 2024 20:19 |
|
|
Claverjoe is correct. Solar electricity is a horribly inefficient way to generate heat. It's on the scale of having to spend hundreds of thousands of dollars to heat the project you are describing. I'm not clear on why solar water heating collectors would not work in your case. Most modern systems aren't really heating the water for end use of having hot water; they are used to heat a glycol-type mixture that is then cycled through heat exchangers for hot water, or through a radiant floor heating system for heating house space. It sounds like you have some of that infrastructure already in place. What's to stop you from using a SWH system to preheat in this case? I honestly don't know SWH system design, a co-worker does though, but I'd think a modest system in the $10k-$20k range would make a huge contribution to your heating needs.
|
|
#
?
Feb 26, 2013 03:05
|
|
|
Yeah, I'm not sure how it would be so difficult to put in another heating loop into an existing system. It should just be a pump to the solar water heaters and then boom, you got it.
|
|
#
?
Feb 26, 2013 05:04
|
|
|
I thought you were referring to individual systems on each building. The hot water system in place also uses a water-glycol mixture, to reduce costs the system is designed to use filtered tap water. Adding another loop is possible, but requires me to redo all of the plumbing. My current pump is nearing capacity. Grundfos pumps aren't cheap. And the area I am referring to doesn't have a lot of enclosed space. The wood stove is an older heatmor 600 CSS. Link: http://www.heatmor.com/commercial-models.php To be quite honest though, If i have a system going from solar to electric to heat, it's inefficiencies don't bother me too much; since there is only upfront cost and maintenance, not fuel. Couldn't I set up a small panel on the roof of the shed to generate enough power to keep a heater coil warm in the water tank, with a control to modulate the two burners (one being the coil, the other being the furnace fans) and dump excess electricity to the ground? This is probably a good time to mention my job is designing and installing temperature controllers for NG/LP burners.
|
|
#
?
Feb 26, 2013 20:06
|
|
|
CBD posted:I thought you were referring to individual systems on each building. I am a scientist and not a plumber (or a petroleum engineer, but I digress), but can you not have a separate loop that has it's own small pump for the water-glycol mix to the tank and let the old pump keep up with the other systems? Or possibly have a separate loop that goes into the tank to heat it, but doesn't share fluid? Yeah solar electric is doable, and if it the difference between being able to install it yourself and having to pay the man, then maybe the cost difference is small, but I don't know without going into spreadsheets and nitty gritty stuff. Isn't there a guy on SA that does solar hot water stuff in the American NorthEast region?
|
|
#
?
Feb 26, 2013 23:25
|
|
|
Claverjoe posted:I am a scientist and not a plumber (or a petroleum engineer, but I digress), but can you not have a separate loop that has it's own small pump for the water-glycol mix to the tank and let the old pump keep up with the other systems? Or possibly have a separate loop that goes into the tank to heat it, but doesn't share fluid? Yeah solar electric is doable, and if it the difference between being able to install it yourself and having to pay the man, then maybe the cost difference is small, but I don't know without going into spreadsheets and nitty gritty stuff. Isn't there a guy on SA that does solar hot water stuff in the American NorthEast region? You mean like a liquid-to-liquid inter cooler type setup? I suppose I could do that. My statement regarding career choice was referring to my ability to setup a control unit for the system.
|
|
#
?
Feb 27, 2013 03:06
|
|
|
CBD posted:Couldn't I set up a small panel on the roof of the shed to generate enough power to keep a heater coil warm in the water tank, with a control to modulate the two burners (one being the coil, the other being the furnace fans) and dump excess electricity to the ground? No. Look at the wattage and time requirements for those loads, and then compare them to what you can expect from a "small panel". Let's use a 140 watt panel(a medium sized panel) as an example. It may cost around $300 or so. On an average sunny day it may collect about 500 usable watthours. A 100 watt(very small) heating element could be run for 5 hours, a 500 watt(still small) element could be run for one hour. What is the wattage of the controller? Is it on 24 hours a day? A 10 watt controller running 24 hours a day would use 240 watthours, or half of what the panel would collect. Adding to the cost, you'll need a battery bank, charge controller, some fusing, and a small inverter if anything is running off of 120 volts AC. You're probably in for $1,000 for the basics and accomplishing little to nothing. Don't even bother considering net-metering, because; there won't be any excess power to sell, it will require even more expensive equipment, and even if you sold all 500 watthours back to the grid, you'd generate about 5 cents in credit for the day. A strange aspect of my job is actually talking people out of buying solar when it is not appropriate for the application. It happens more than you'd expect. We get to use our discretion at work and all of us use the rule, "I won't recommend anything that I would not use myself". It's a very good rule, because it earns trust and often leads to realistic expectations and sales for other applications.
|
|
#
?
Feb 27, 2013 03:17
|
|
|
CBD posted:You mean like a liquid-to-liquid inter cooler type setup? I suppose I could do that. My statement regarding career choice was referring to my ability to setup a control unit for the system. Ohhh, okay. But yeah, solar water heater assist to your tank would almost certainly be the application you would want, which is nice since solar hot water is much, much cheaper than solar electric.
|
|
#
?
Feb 27, 2013 05:14
|
|
|
CBD posted:You mean like a liquid-to-liquid inter cooler type setup? I suppose I could do that. My statement regarding career choice was referring to my ability to setup a control unit for the system. This would be ideal. You've got the buildings already plumbed with the wood-furnace system. This would just involve adding the liquid-liquid heat exchanger between the solar and wood systems. Install it before the furnace so that it never ends up acting as a radiator for your wood-heated water. On days with good sun, you're harvesting heat without any photovoltaic losses or having to pay $pain for solar panels. Even on days with crap solar heating, it's still acting as a pre-heater for the furnace and cutting down on the amount of wood needed. The other advantage of the heat exchanger is that if the solar system leaks due to hail or chupacabras, it doesn't take down the existing system.
|
|
#
?
Feb 27, 2013 17:04
|
|
|
Do you anticipate (or have you noticed in your time working in the solar business) prices for panels, wiring, installation, etc going down to a level which would make solar energy feasible for many or most middle-class homeowners? I would love to have a grid-tie system on my hypothetical future home, but I live and probably always will live in a very low cost-of-living area, where I could buy an entire house for the price of a solar panel array.
|
|
#
?
Feb 27, 2013 17:14
|
|
|
Pricing on solar panels is as low as it has ever been. The 140 watt panel I mentioned above for $300, was about $650 just 4 years ago, and around $800 8 years ago. Larger panels, in the 200+ watt range, have a a much better price per watt, approaching $1 per watt. Can it go lower? Possibly. Maybe down to 80 cents or so per watt. At a certain point I'd expect prices simply cannot go noticably lower. Not sure where that point is. That's all with crystalline silicon panels. I'm sure some of the newer technologies will become feasible for mass market, retail consumption in the next 10-30 years. Panels with higher efficiencies per square foot based on newer tech. They'll start off at high cost, and the whole price battle will start over for the following 20 years. At that time the traditional crystalline panel will probably be at the cheapest price possible, as the low cost alternative. Lead acid batteries will only get more expensive. Scaling up with normal inflation, and scarcity of cheap resources. Newer tech batteries like lithium will be interesting. When the electronics to properly utilize them in RE systems becomes common and widespread, the cost of the energy storage component of a system will go down. I'd have to guess 5-15 years for lithium specifically. Things like cabling, fusing, and breakers will all ride along at current costs, adjusted for inflation. Nothing to change there. The electronics, such as the charge controllers and inverters have lots of room for innovation. There will probably be a lot of more expensive, but also much more useful and highly functional products coming to the market on a yearly basis. MPPT controllers are a great example. More expensive controller, but also has the ability to harvest more usable power off a fixed array size. Picking and choosing those products will be a function of researching what's available and proper system design practices.
|
|
#
?
Feb 28, 2013 17:01
|
|
|
What would I need for a small hobby-ist setup? I'd like to setup a sort of battery bank that will act as an extra power supply for a computer for say, several hours. By calculations, a single deep cycle battery with 225 Amp hours could handle 500 watts for 2 hours. I'd like to recharge that battery during the day with a 100w panel (maybe more but I want to start on the cheaper side). And I would also need a power inverter ? The only thing I don't understand is the exact setup and perhaps what are good brands/parts to buy for this. Would I be able to just hook the battery/battery bank to the power inverter, then hook the solar power directly to the batteries or what? EDIT: I'm open to a 100w wind turbine as well if those exist. Sylink fucked around with this message at 06:02 on Apr 25, 2013 |
|
#
?
Apr 25, 2013 05:33
|
|
|
Hey look a post! The basics you will need will be a panel, a charge controller, a battery, an inverter, and some fusing for basic safety. I'll leave cabling, mounts, meters, etc out of the mix, for the sake of discussing the basics. Does your rig really pull 500 watts? My old HP tower with graphics card and two monitors only used about 250 watts. My one year old Dell laptop only uses less than 40 watts. A single 12 volt battery at 225AH has a total capacity of 2,700 watt hours. A 500 watt load for 2 hours will use 1000 watt hours, or a discharge of 37% per day(no sun, no charging). A bit on the heavy side(I would normally recommend a discharge of 15% as a sizing goal), meaning a shorter battery life expectancy, but not totally unreasonable. I would shoot for a panel, or total array size of at least 250-300 watts to collect 1000 usable watthours per day. More if you are further north, or in cloudier climates like Minnesota. 200 minimum though, even if you are in New Mexico. Two twelve volt panels wired in parallel, with a traditional 3-stage PWM charge controller that can handle at least 20 amps should be good. A Xantrex C-35 is the old standby at a decent price. Put a fuse of about 30 amps in the positive leg between it and the battery. An alternative would be a single large panel(60 cell module, produces about 30 volts) with an MPPT controller. Much cheaper price on the panel, but a more expensive controller. Its kind of a wash at the smaller array size we are talking about. If the battery is a flooded lead acid type(the kind with caps that you can add distilled water to) then it will give off gasses when charging. Make sure it is well ventilated naturally or use an active fan to draw the gasses away, as they are explosive and corrosive. If you smell sulfur/rotten eggs when its charging, then the gasses are building up to bad levels. I'd recommend a pure sine wave inverter since you are running expensive electronics off of it. Obviously something 12 volt based that can handle more than 500 watts continuous. Samlex has some inverter only units in the 600 and 1000 watt range. I honestly would not recommend them for full time residential applications, but they are fine for hobbyist purposes, and also a decent price. Follow the manual's instruction for cable sizing and hook it straight to the battery bank, through a proper fuse in the positive leg as well. Check the manual for the fuse size, something in the 100-200 amp range will be appropriate on smaller inverters. The Samlex units have the advantage of having outlets built right into the front of them, so you can plug the power strip or wall cube of your computer straight into them. This scheme has no backup method of charging the batteries for extended periods of overcast weather. Consider a more expensive inverter/charger(built-in battery charger) or a stand alone charger; either of which will be used with utility power or a generator. Or, expect to not use the system when the sun doesn't come out and the batteries start to approach the 50% state of charge. Or, become a battery murderer and discharge them more than that on a regular basis. Skip all small scale wind turbines. Don't even bother for something like this.
|
|
#
?
Apr 26, 2013 02:24
|
|
|
Thanks for the info! What batteries and panel manufacturers do you recommend? And if I wanted to set it up to use my home grid when their isn't sufficient charge/sun and have it switch automatically how difficult is that? Or at that point would it be easier to talk to a utility company and see how about adding it into the entire house grid?
|
|
#
?
May 21, 2013 00:23
|
|
|
How long until high efficiency solar panels become available for consumer use? I've seen a few things that have high eff., but are just prototypes from like IBM and so on. Are there any companies that will sell sun-tracking parabolic reflector solar dishes to consumers? From what little gleaming I've done, parabolic reflectors used on tiny solar cells are more efficient than the normal flat solar panels, but generate enormous heat.
|
|
#
?
May 23, 2013 02:10
|
|
|
SlayVus posted:How long until high efficiency solar panels become available for consumer use? I've seen a few things that have high eff., but are just prototypes from like IBM and so on. Are there any companies that will sell sun-tracking parabolic reflector solar dishes to consumers? From what little gleaming I've done, parabolic reflectors used on tiny solar cells are more efficient than the normal flat solar panels, but generate enormous heat. Probably never. Okay, let me give you the long answer, physical scarcity of the necessary elements limits those super high efficiency panels to places with major concentrators and cooling systems (good luck getting permitting for that in your backyard) or places where price isn't a secondary or even a tertiary concern (like in space). Consumer solar cells hit the sweet spot of cost to produce versus efficiency, and most tricks used to make high efficiency solar cells cheaper will be applied to the standard silicon solar cells as well, keeping them far in the lead as a cost-to-performance race.
|
|
#
?
May 23, 2013 18:16
|
|
|
Claverjoe posted:Probably never. Okay, let me give you the long answer, physical scarcity of the necessary elements limits those super high efficiency panels to places with major concentrators and cooling systems (good luck getting permitting for that in your backyard) or places where price isn't a secondary or even a tertiary concern (like in space). Scarcity of elements is kind of an overblown issue. You probably are referring to GaAs solar cells and are referring to the gallium and indium used in those cells. I have been told that gallium and indium aren't even mined for--all of the production of those elements is from byproducts of aluminum mining and refining. Since those materials are great absorbers, you don't need a lot of them either. You are right though that it remains to be seen whether they can be cost competitive with silicon even though they do have an efficiency advantage. There is a company which makes flexible flat-plate GaAs cells which is trying to knock the materials cost out of those solar cells by exfoliating the solar cell from the native substrate, attaching it to plastic, and reusing the GaAs substrate. This method also has the benefit of improving performance as well. silence_kit fucked around with this message at 16:03 on May 24, 2013 |
|
#
?
May 24, 2013 15:24
|
|
|
The thing is, for most terrestrial applications, area and weight aren't big concerns. If you can have a cheap-rear end per square foot solar panel taking up your entire roof in order to get enough sunlight or a very fancy solar panel taking up one square meter, why choose one over the other if the total system cost is similar?
|
|
#
?
May 24, 2013 18:07
|
|
|
Frozen Horse posted:The thing is, for most terrestrial applications, area and weight aren't big concerns. If you can have a cheap-rear end per square foot solar panel taking up your entire roof in order to get enough sunlight or a very fancy solar panel taking up one square meter, why choose one over the other if the total system cost is similar? Note that I am not talking about concentrator cells. Concentrator cells have additional complexity and safety issues associated with them. They aren't for homes. Are you seriously asking this question? The hypothetical 1m^2 solar panel (which is not possible by the way because unconcentrated sunlight is not that strong) that could power an entire home would be way better. In heavily populated areas, solar power would make more sense. Real estate isn't always cheap. silence_kit fucked around with this message at 20:06 on May 24, 2013 |
|
#
?
May 24, 2013 20:01
|
|
|
silence_kit posted:Scarcity of elements is kind of an overblown issue. You probably are referring to GaAs solar cells and are referring to the gallium and indium used in those cells. I have been told that gallium and indium aren't even mined for--all of the production of those elements is from byproducts of aluminum mining and refining. Scarcity of elements is a major issue if you actually want to deploy a significant number of solar cells. Once you tap out the waste stream material, you hit a hell of a brick wall in terms of procuring more. Flexible flat plate Ga-As is not a triple junction InGaP/InGaAs/Ge solar cell with anti-reflection coatings either. Ga-As is NOT a 33+% high efficiency solar cell. Cell record efficiency is, what, 18.4% efficient, as the 28%+ efficiency was with a cell less than 1 cm2, which is the threshold for an official record by academia/Martin Green, though it still gets published.
|
|
#
?
May 24, 2013 21:31
|
|
|
silence_kit posted:Scarcity of elements is kind of an overblown issue. You probably are referring to GaAs solar cells and are referring to the gallium and indium used in those cells. I have been told that gallium and indium aren't even mined for--all of the production of those elements is from byproducts of aluminum mining and refining. No, the issue of scarcity is more for CdTe. 10 GW of CdTe solar cells (about 1 large power plant) would conceivably use the entire world's supply of tellurium. A lot of metals aren't mined for. You can't necessarily find economic concentrations of certain elements. Platinum is a by-product of copper and nickle mining, and there are lots of useful and expensive applications for platinum too.
|
|
#
?
May 24, 2013 21:31
|
|
|
Claverjoe posted:Scarcity of elements is a major issue if you actually want to deploy a significant number of solar cells. Once you tap out the waste stream material, you hit a hell of a brick wall in terms of procuring more. I am not really convinced that indium or gallium are particularly scarce. There at least seems to be enough indium for the indium tin oxide in LCD screens. LCD screens are cheap and plentiful. Claverjoe posted:Flexible flat plate Ga-As is not a triple junction InGaP/InGaAs/Ge solar cell with anti-reflection coatings either. Ga-As is NOT a 33+% high efficiency solar cell. Cell record efficiency is, what, 18.4% efficient, as the 28%+ efficiency was with a cell less than 1 cm2, which is the threshold for an official record by academia/Martin Green, though it still gets published. The module efficiency record is 23.5%. The company's technique could work on multiple junction solar cells as well, as long as there isn't too much aluminum in the solar cell. silence_kit fucked around with this message at 01:20 on May 25, 2013 |
|
#
?
May 24, 2013 23:39
|
|
|
As I understand it, if you install home solar panels in my area, you're forced to choose if you want to be on or off the grid with them. If you're off the grid, you feed the power directly back into your house, but you can't sell excess back to the power company. If you're on the grid, you can sell excess back to the power company, but if the power goes out, you don't get to use your own electricity from the solar panels. I have to think that this is a false dilemma. Is it really not possible to install them on the grid, then have an electrician put in a line to your home power, and a switch so that the owner can switch between the two options as desired?
|
|
#
?
May 25, 2013 16:41
|
|
|
I don't know about regulations where you live, but it is certainly possible. Here are two posts about a home setup where he can run off grid or take utility power if necessary.
|
|
#
?
May 26, 2013 02:45
|
|
|
So my boss wants to put in an electric vehicle charging station. He figures that it'd be a neato conversation piece if he could tell people it was also hooked up via Solar. So I have 30 square feet of space in Atlanta, the vehicle the thing is gonna be charging requires 24 Kilowatt hours per charge and my boss figures it'd only be used 10 times a month or so, Is this feasible? How expensive would this run? I assume in this case it'd be about the battery rather than the panels itself?
|
|
#
?
Jun 3, 2013 15:55
|
|
|
KildarX posted:So my boss wants to put in an electric vehicle charging station. He figures that it'd be a neato conversation piece if he could tell people it was also hooked up via Solar. So I have 30 square feet of space in Atlanta, the vehicle the thing is gonna be charging requires 24 Kilowatt hours per charge and my boss figures it'd only be used 10 times a month or so, Is this feasible? How expensive would this run? I assume in this case it'd be about the battery rather than the panels itself? Solar plus grid with net metering is much better of an idea than a battery bank for this. When someone is charging and the sun is shining, the power is coming from the panels. When someone is charging at night, the power is coming from the grid. When someone isn't charging, the power is going to someone running their AC elsewhere in Atlanta thus reducing the load and fuel consumption on the power plant.
|
|
#
?
Jun 3, 2013 17:30
|
|
|
Sylink: Consider a set of Kyocera 140 watt panels. Not the cheapest, but meets the criteria of: long-standing manufacturer, good warranty, solid panel. Just a suggestion, as price is always important. For batteries, a Trojan T105 is the old standby for the same reasons. Probably one of the most common batteries in the US. They are 6 volts, so pairs of 2 for a 12 volt configuration(2 or 4 or 6 depending on how much capacity you want). If you want to use grid power as a backup to keep the batteries topped off when the sun is not around, either get a stand-alone battery charger like the Iota5512(bulk charger only, turns into a 3 stage charger if you buy the IQ4 accessory) for quality at a lower cost. Or invest in a nicer inverter; like a Magnum MS2012(get the RC50 control interface as well, should be free with the inverter purchase). The utility company doesn't care when you pull power *off* the grid for use with battery charging. It's essentially like any other household load. They only care when you try to put power *on* to the grid, when selling back. No point in considering selling back for small scale hobby systems. So no need to buy equipment with that type of feature set. Peristalsis: Some more precise configuration definitions/possibilities: Pure Off-grid: battery based, no utility power at all. Off-grid with utility backup: battery based, mostly pure off-grid, but uses grid power as a backup only(instead of, or along side with, a generator). NO selling to the grid. Pure Grid-tie: No batteries. Power generated offsets the household's use, excess is sold back to the grid. Does nothing during utility outages. Grid-tie with battery backup: Pure grid-tie, but with a battery bank for outages. The battery bank is typically smaller to only accommodate a few critical loads such as a refrigerator, a few lights, and a computer or tv. The battery bank is usually NOT sized to backup the whole house(too expensive/ridiculous). The house functions as a grid-tie house, as well as keeping the battery bank topped off and ready for utility outages. When the power does go out, the equipment used will isolate the household from the grid connection and run the system as an off-grid battery based house for the time being. Two considerations: net-metering, or "selling back" excess power is only worth it if you have a decent size array of a couple thousand watts or (much)more. Don't bother if you only want a few hundred watts of solar panel. Also, choosing a grid-tie system with battery backup should be decided on at day one. It will be much cheaper to implement. An existing grid-tie system *can* be retrofitted(and doing so is in high demand nowadays) to add in a battery bank at a later date, but the overall system cost ends up be at least $3000-8000 more. Plan from day one if you want a grid-tie with battery backup, and save yourself piles of money. KildarX: A full charge of 24 kWhrs every three days would average 8 kwh per day for system sizing purposes. You would want 2000 watts of panels or more(depending on where you live) to collect that much power in one day. For typical battery bank sizing, 8 kwh per day would require a 48 kwh battery bank. A 48 volt battery bank would be appropriate with 1000 amphour capacity. A big bank, using industrial sized Surettes, or HuPs, or something similar would be wanted. Overall though, an off-grid system in the range of $30,000 - $40,000 is what would be needed. If that is what he wants, its doable, just not cheap. Science/materials Good stuff! Not my forte. I enjoy reading it as well though.
|
|
#
?
Jun 11, 2013 04:45
|
|
|
silence_kit posted:I am not really convinced that indium or gallium are particularly scarce. There at least seems to be enough indium for the indium tin oxide in LCD screens. LCD screens are cheap and plentiful. quote:The module efficiency record is 23.5%. The company's technique could work on multiple junction solar cells as well, as long as there isn't too much aluminum in the solar cell. Multijunction solar cells need the absorbing parts of their layers to 1) have complimentary absorption profiles so they don't interfere with each other and 2) have the physical absorption layers match up the currents produced from each junction, since the voltages can't be easily matched without playing voltage boost/buck games, which eat into effect collection efficiencies as well, so what is the absorption profile, and what are the physical thickness of the proposed multijunction solar cell? You can't just slap on on top of another and call it a day. Look I'm not saying it is impossible, but I am saying that the question is probably a bit more complex than you think.
|
|
#
?
Jun 11, 2013 07:19
|
|
|
Thanks again for the reply! Gonna start budgeting things out. EDIT: Do you have a preference for pure sine inverters that are close to 1000W? I know you recommended one already but that was pretty expensive and I'm not ready for the big boys yet. Will I be safe just picking something that isn't rated horribly and is pure sine? I just searched by reviews on amazon and found a few, for instance - http://www.amazon.com/Sunforce-1124...e+wave+inverter Sylink fucked around with this message at 09:35 on Jun 11, 2013 |
|
#
?
Jun 11, 2013 07:36
|
|
|
I can't say that I know of the Sunforce brand, so I can't comment either way. In general, "You get what you pay for" is really, really true when in comes to solar electric equipment. I'm fine with less expensive gear for hobbyist, temporary usage, etc, applications. I wouldn't choose it for whole house, mission critical situations though. I'd expect 5 years of life on a 1000 watt sine at that price point. If you get more, then all the better.
|
|
#
?
Jun 11, 2013 14:04
|
|
|
Claverjoe posted:This is a hand-wave, and not any kind of real analysis. How many tons are needed for how many GW, and how many tons are available at current price point? What happens to the price curve after all waste streams are tapped out? Well, you are the guy who started it by claiming that gallium and indium are so scarce, why don't you actually show that this is true instead of pulling assertions out of your rear end? According to this report http://pubs.usgs.gov/circ/1365/Circ1365.pdf from the United States Geological Survey, there doesn't appear to be a scarcity problem for indium and gallium. It even noted that most of the gallium extracted from aluminum refining is actually thrown away as waste. Claverjoe posted:Multijunction solar cells need the absorbing parts of their layers to 1) have complimentary absorption profiles so they don't interfere with each other and 2) have the physical absorption layers match up the currents produced from each junction, since the voltages can't be easily matched without playing voltage boost/buck games, which eat into effect collection efficiencies as well, so what is the absorption profile, and what are the physical thickness of the proposed multijunction solar cell? You can't just slap on on top of another and call it a day. Look I'm not saying it is impossible, but I am saying that the question is probably a bit more complex than you think. I understand what a tandem solar cell is and some of the design issues that are encountered in designing a tandem solar cell. What I am claiming is that you could pretty much take an existing gallium arsenide tandem solar cell design, add a few layers, and process it in the way that the company does by removing the substrate, and you could get a flexible tandem cell. The main issue that I pointed out in a previous email is that the process of removing the substrate prevents you from using aluminum-rich alloys of gallium arsenide and aluminum arsenide.
|
|
#
?
Jun 11, 2013 21:46
|
|
|
silence_kit posted:Well, you are the guy who started it by claiming that gallium and indium are so scarce, why don't you actually show that this is true instead of pulling assertions out of your rear end? Uhh, that link was for a case of refining capacity and current amounts taken up by the world's refining capacity, not world reserves. You might want to read the paper you are touting more closely. Try this: http://minerals.usgs.gov/minerals/pubs/commodity/indium/indiumcs07.pdf For Indium: ~ 6,000 metric tons, at current price point, refining from zinc ores. Now, from your article, they go with roughly 90 tons of Indium for 8760 GWh of annual capacity, ignoring other material inputs (at idealized conditions, oops!). which I don't personally agree with, but fine, whatever, we'll use it, which leaves us at an absolute maximum of 0.584 PWh/year production, using every scrap of possible indium, and assuming there are no more LCD TVs, or most other electronic devices which people are willing to pay more for than solar panels, which sounds like a joke in poor taste. Total world energy consumption: 150 Pwh/year, and rising. Now that being said, that 90 tons is only about 17% of what is being used every year, so we can assume that really, of that 6000 metric tons of recoverable material, we can do a generous estimate saying about 25% of it will find its way to solar cells, assuming that they are wonderfully successful, everybody loves solar cells and suddenly hates smartphones and TVs, rah rah rah, so that 0.584 PWh/year becomes something more like 0.146 Pwh/year. Now there are more technically recoverable sources of Indium, but if the process gets too expensive, the entire reason for using indium based thin film solar cells (cost advantage) evaporates, and it becomes a worthless exercise of techno-masturbation. Otherwise known as "who gives a poo poo". Now for Gallium, it says straight up that the refining process of Gallium is dog poo poo for throughput because the crystals are too expensive to make in your paper. How much are your favorite solar cells selling for *right now*? Hell, even Alta Devices admits up front that silicon solar cells are cheaper in a cost/watt evaluation, so their method has some serious processing costs at the very least. Look enthusiasm is all well and good, but even the company you are boosting is saying they aren't competitive right now, and you want to stack on more processing costs for diminishing returns? Nuts to that idea. I did my Phd. dissertation in materials science on the topic of solar cells, I spent five years of my life reading on solar cells from a researcher's perspective. I'd like to think I know a little bit after spending half a decade on it. The rule for *any* mass deployed Solar cell Technology is going to be this: It must make the element 26 limbo, because anything above that elemental number is near guaranteed to be insufficient in amount. It's why there is so much push for plastic solar cells and the like, even thought their efficiency is poo poo, making tons of plastic is never going to be a problem. The Dipshit fucked around with this message at 02:38 on Jun 12, 2013 |
|
#
?
Jun 11, 2013 22:36
|
|
|
Claverjoe posted:Now for Gallium, it says straight up that the refining process of Gallium is dog poo poo for throughput because the crystals are too expensive to make in your paper. How much are your favorite solar cells selling for *right now*? Yes, I know that gallium arsenide wafers are expensive--however, you don't need the wafers to be incorporated into the solar cell and in principle, you could reuse the wafers to make other cells. They aren't my favorite solar cells--I am just challenging your assertion that GaAs will forever be doomed for high cost and pointing out that there are engineering approaches to lower the cost. In my first post in this thread I noted that it remains to be seen whether it can be cost-competitive with silicon. Claverjoe posted:I did my Phd. dissertation in materials science on the topic of solar cells, I spent five years of my life reading on solar cells from a researcher's perspective. I'd like to think I know a little bit after spending half a decade on it. So? Loudly proclaiming that you are an expert in something doesn't replace actually using that expertise. silence_kit fucked around with this message at 05:14 on Jun 12, 2013 |
|
#
?
Jun 12, 2013 05:11
|
|
|
Okay, I admit, I've been a bit caustic due to irritation at your inability to read your own documents you bring to the table, but here are some honest answers, now that I've laughed it off.silence_kit posted:Yes, I know that gallium arsenide wafers are expensive--however, you don't need the wafers to be incorporated into the solar cell and in principle, you could reuse the wafers to make other cells. They aren't my favorite solar cells--I am just challenging your assertion that GaAs will forever be doomed for high cost and pointing out that there are engineering approaches to lower the cost. In my first post in this thread I noted that it remains to be seen whether it can be cost-competitive with silicon. Okay, apparently we have found the disconnect here, you need the wafers to make the cells. The technology that these people are using are making very thin slices from the ingot using not a bladed saw, which is the standard operating procedure for making Si-cells, but a liftoff technique that makes very thin wafers. The wafer is used up ~10x slower than the current bladed saw method, but you can't make a GaAs solar cell without the GaAs, and the Si solar cells are *still* so much cheaper due to relative rarity of the material compared to Si, the thermodynamic cost of making the GaAs, and due to the fact that courtesy of the computer industry, Si is *the* most well understood element known to humanity. There is no workaround to the thermodynamic cost of making GaAs ingots, and the technically recoverable reserves of Ga are pretty dismal, since there isn't really a gallium ore hanging around in the earth's crust. (http://minerals.usgs.gov/minerals/pubs/commodity/gallium/mcs-2012-galli.pdf) Challenge all you want, but it is physically impossible for Ga to do any serious heavy lifting in terms of solar cell production, even with all the neat engineering tricks possible. silence_kit posted:So? Loudly proclaiming that you are an expert in something doesn't replace actually using that expertise. I sincerely dislike having to repeat basic, well-known information to people who can't read and draw really silly conclusions from their illiteracy. It strikes that same emotional chord who think that lizard-people are secretly ruling the world. 
|
|
#
?
Jun 12, 2013 06:41
|
|
|
Claverjoe posted:Okay, apparently we have found the disconnect here, you need the wafers to make the cells. The technology that these people are using are making very thin slices from the ingot using not a bladed saw, which is the standard operating procedure for making Si-cells, but a liftoff technique that makes very thin wafers. The wafer is used up ~10x slower than the current bladed saw method, I don't understand what you mean here. I kind of suspect that you don't understand the company's technology. The solar cells aren't actually made of wafer material--the solar cells are made of material grown epitaxially on the wafers. The wafer isn't used up at all, so at least in principle, you can reuse the wafers as many times as you want. Claverjoe posted:but you can't make a GaAs solar cell without the GaAs, and the Si solar cells are *still* so much cheaper due to relative rarity of the material compared to Si, the thermodynamic cost of making the GaAs, and due to the fact that courtesy of the computer industry, Si is *the* most well understood element known to humanity. Thermodynamic cost of making GaAs? What do you mean by this? Silicon is a much higher temperature material than gallium arsenide, so naively I would think that the energy cost of making a gallium arsenide crystal is less than that of a silicon crystal. quote:the technically recoverable reserves of Ga are pretty dismal, since there isn't really a gallium ore hanging around in the earth's crust. (http://minerals.usgs.gov/minerals/pubs/commodity/gallium/mcs-2012-galli.pdf) Challenge all you want, but it is physically impossible for Ga to do any serious heavy lifting in terms of solar cell production, even with all the neat engineering tricks possible. Pointing this out isn't the same thing as showing that there is a material scarcity problem for gallium arsenide solar cells.
|
|
#
?
Jun 12, 2013 07:52
|
|
|
silence_kit posted:I don't understand what you mean here. I kind of suspect that you don't understand the company's technology. The solar cells aren't actually made of wafer material--the solar cells are made of material grown epitaxially on the wafers. The wafer isn't used up at all, so at least in principle, you can reuse the wafers as many times as you want. The solar cell maintains a material cost. The GaAs has to be put into the solar cell, which can't be used for other solar cells. I thought that you were thinking that there was a zero material cost for each device. I believe that when you used the word wafer, a more appropriate word would be "anvil" or "wafer substrate" and is the cause of the miscommunication. It is better to think of the "wafer" as you use it as a piece of equipment to make the solar cell rather than the actual solar cell component itself. Epitaxial growth is a (usually chemical vapor) deposition process which takes up usage of a material to a thickness of a few micrometers, like around 10 um or so. Since a regular solar cell wafer is on the order of 100 um thick these days, there isn't a compelling amount of difference in the material to make up the difference in availability. It's nice and all they have a good liftoff technique to transfer the grown material to something else, but that doesn't change the material cost. Amorphous silicon solar cells are made in a much similar way (not called epitaxial growth, since that means a growth with a crystalline orientation), and it isn't something new under the sun. quote:Thermodynamic cost of making GaAs? What do you mean by this? Silicon is a much higher temperature material than gallium arsenide, so naively I would think that the energy cost of making a gallium arsenide crystal is less than that of a silicon crystal. When your raw ore comes in 10-50 parts per million, there is a significant cost in energy, time, and money to go from 0.00001% to 100% material that you can work with which is before you even get to think about making a crystalline solid. In essence, purification from such levels is a bitch on wheels. The "raw ore" for silicon comes as, well, sand and quartz, which is something like 20% Si, and is the second most physically abundant element in the Earth, behind oxygen. quote:Pointing this out isn't the same thing as showing that there is a material scarcity problem for gallium arsenide solar cells. There IS a materials scarcity of available GaAs solar cells. Just because X amount of element exists, if it is not recoverable in any sensible fashion, it can't be considered available. As an example, right now, there is probably a diamond that is roughly the size of earth orbiting another star (http://abcnews.go.com/blogs/technology/2012/10/a-planet-made-of-diamond-twice-the-size-of-earth/), and just because it exists, doesn't mean it has any influence on the market price of diamonds here, because we can't reach it in any sensible fashion to mine it for use. As another example, oil wells have two values associated with them, the oil in the formation, and the amount of oil that is recoverable, which can get up pretty high these days for conventional fields, but it isn't 100% of the oil in place, which nobody pays for because it is still in the rock. The Dipshit fucked around with this message at 14:01 on Jun 12, 2013 |
|
#
?
Jun 12, 2013 13:58
|
|
|
Claverjoe posted:Epitaxial growth is a (usually chemical vapor) deposition process which takes up usage of a material to a thickness of a few micrometers, like around 10 um or so. Since a regular solar cell wafer is on the order of 100 um thick these days, there isn't a compelling amount of difference in the material to make up the difference in availability. The gallium arsenide cells need only to be like one to three microns thick to get all of the current. Silicon solar wafers aren't 100um thick-that's like the theoretical minimum thickness of a silicon cell. They are like 200-300 microns thick, and that's not counting the material wasted in sawing and polishing. I have been told that in the process of turning boules into wafers more material gets wasted in sawing and polishing than what actually goes into the wafer. Assuming that the wafers can be reused enough times to make wafer material usage negligible to the epitaxial material usage that's at least a two orders of magnitude reduction in material use that I am pointing out here. Claverjoe posted:It's nice and all they have a good liftoff technique to transfer the grown material to something else, but that doesn't change the material cost. No it does! You aren't understanding the claim. If you can reuse the wafer to make more cells you can amortize the wafer cost over many cells until it becomes negligible. Claverjoe posted:Amorphous silicon solar cells are made in a much similar way (not called epitaxial growth, since that means a growth with a crystalline orientation), and it isn't something new under the sun. I am not claiming that this company invented epitaxy. The epitaxy technology that this company uses for gallium arsenide is thirty years old, if not older. It is commonly used in LED and diode laser manufacturing. Claverjoe posted:When your raw ore comes in 10-50 parts per million, there is a significant cost in energy, time, and money to go from 0.00001% to 100% material that you can work with which is before you even get to think about making a crystalline solid. In essence, purification from such levels is a bitch on wheels. The "raw ore" for silicon comes as, well, sand and quartz, which is something like 20% Si, and is the second most physically abundant element in the Earth, behind oxygen. Silicon needs to be purified as well. I guess the cost of refining gallium is sort of built in to the aluminum refining process. quote:There IS a materials scarcity of available GaAs solar cells. Just because X amount of element exists, if it is not recoverable in any sensible fashion, it can't be considered available. I question whether you need a lot of material in the first place to make the cells. The waste stream from aluminum mining may be enough.
|
|
#
?
Jun 12, 2013 16:42
|
|
|
No, just no.silence_kit posted:The gallium arsenide cells need only to be like one to three microns thick to get all of the current. Silicon solar wafers aren't 100um thick-that's like the theoretical minimum thickness of a silicon cell. They are like 200-300 microns thick, and that's not counting the material wasted in sawing and polishing. I have been told that in the process of turning boules into wafers more material gets wasted in sawing and polishing than what actually goes into the wafer. Hell, here's a blog post on thin Si, since I'm wagering you don't have access to journals: http://green.blogs.nytimes.com/2012/03/13/slicing-silicon-thinner-to-cut-the-price-of-solar-cells/ They won't be as efficient at the theoretical best thickness (around 50 um or so, as I recall), but who cares if they work and are cheap. quote:No it does! You aren't understanding the claim. If you can reuse the wafer to make more cells you can amortize the wafer cost over many cells until it becomes negligible. quote:Silicon needs to be purified as well. I guess the cost of refining gallium is sort of built in to the aluminum refining process. The point is that silicon takes a very short road to purify compared to Ga. 0.2 starting point is much better than 0.00000001. Sure you are making use of a waste stream, but you are going from 0.00000001 to 0.0000001 from that separation from the waste stream into starting its own chain of working upon making Ga. quote:I question whether you need a lot of material in the first place to make the cells. The waste stream from aluminum mining may be enough. But wait, let's assume that instead of what is possible, let's go with 100% recovery of Ga from all world ore, with a 10 ppm concentration, to make the math easy. 29,000,000,000 tons, as a world reserve, giving us 29*10000, 290000 tons of Ga available from all the worlds aluminum that we have left, giving us a nice, respectable 846.8 PWh/year maximum possible total world energy production from Ga cells, assuming that the USGS paper is off by a factor of 10 for the Ga input (!), assuming that we mine up every single source of aluminum that humanity knows about (!!!), assuming a near impossible amount of Ga recovered from the ore, and assuming that every single ounce of Ga goes to making solar cells. So I guess if you want to handwave literally everything away, you could make a case. The Dipshit fucked around with this message at 17:34 on Jun 12, 2013 |
|
#
?
Jun 12, 2013 17:18
|
|
|
Claverjoe posted:2 orders of magnitude of mass is pissing in the wind, you need 5 or 6 orders of magnitude to compensate for the rarity/concentrations found in the wild. You are assuming that all sand gets turned into silicon or needs to get turned into silicon to make an appreciable amount of solar cells. I doubt that this is true. quote:Your information on the thickness of a solar cell wafer is about a decade out of date, standard ones are going with 120-150, and ultrathin Si solar cells are also quite possible. Do you have a source for this? I am going off of a recent presentation given by the founder of SunPower, where he gave 250-300um as normal thicknesses. quote:I mean, generally as a rule of thumb, any technique used on one crystalline solid can have a translation to another crystalline solid. This is not true in this case. The technique that the gallium arsenide company uses leverages the facts that 1) hetero-epitaxy is much more mature technology and is richer for the III-V semiconductors over silicon (this is actually a lot of the reason why gallium arsenide cells are higher performance than silicon cells) and 2) that there is a huge difference in solubility of aluminum arsenide and gallium arsenide in hydrofluoric acid. There isn't a way to do what the gallium arsenide company is doing in silicon. There may be other ways to achieve thin silicon wafers without having to waste a lot of silicon though. Claverjoe posted:Hell, here's a blog post on thin Si, since I'm wagering you don't have access to journals: That company went out of business, although I think a lot of other silicon solar cell companies are looking into techniques to make thinner silicon solar cells without wasting so much material. It is a good idea. One thing that I don't understand is how you can process these silicon cells which are exfoliated from the wafer (the technology that Twin Creeks tried to sell is similar to the highest quality way to make silicon-on-insulator wafers). Doping and oxidizing silicon traditionally are high temperature processes, and attaching the thin, fragile silicon films onto glass or plastic sort of prevents you from doing those processes after the fact. The gallium arsenide company puts the dopants in during the growth and the processing of the cells is simple and at low temperatures. This is pretty normal for III-V device processing. I would also like to say that in this thread you've said multiple wrong things about solar cells, so I would knock it off with the condescending know-it-all crap. Claverjoe posted:They won't be as efficient at the theoretical best thickness (around 50 um or so, as I recall), but who cares if they work and are cheap. Well, at least in silicon, half of the cost of solar cells is independent of their manufacturing cost, and improving efficiency allows you to recoup those costs in addition to the manufacturing costs so you can't just declare efficiency to be irrelevant. Obviously you also need higher areas of solar cells to get the same power output, so you aren't really saving on material cost per watt by making your cells too thin to be able to get all of the current that they could get from the sun. Claverjoe posted:Uhh, that's the point of having a machine tool. More and more I'm convinced you are missing something critical on this subject. The wafer is a substrate which epitaxy is performed upon. The epitaxial layer is then lifted off, and that lifted off layer is the solar cell. The wafer is a place to deposit upon, similar to being a mold. Declaring that you can reuse a mold and that it is so amazing is very odd, to say the least. I understand how it works. Spare me the lecture please. It is unusual in that all other III-V devices (okay, let's ignore the blue LED) are made on expensive III-V substrates. All projections on cost, material availability assume that you are attached to a substrate. The substrate is over 100x thicker than what you need to make a solar cell. Claverjoe posted:No. Let's go with your original paper you misread. Presuming that we are going to ignore all other applications of Ga in, well anything, we get 111 metric tons of production each year from tailing and recycling. Lets also assume that the assumption of 30 metric tons per 8760 GWh of production is off by a factor of 10 and that you only need 3 metric tons per 8760 GWh, due to this "ground breaking, paradigm shifting" technology. This leads to 0.324 PWh/year of production, assuming unbelievable levels of production. This is less than the yearly increase in total world demand. It is a rounding error. There is not enough economically recoverable material to work with. Technically, you are misreading the paper--the 30 tons number is for a low-efficiency CIGS technology. The 111 tons number that you cited is also the amount of gallium that they use, not that amount that is available. The paper notes that there is way more gallium available from aluminum refining that people currently don't have a use for and don't bother to use. Also, assuming that all gallium goes into making III-V semiconductors is not that unreasonable of an assumption--gallium isn't used that much. It is very possible that the factor of ten that you donate to the calculation may not be generous enough. There are a lot of uncertainties here. quote:But wait, let's assume that instead of what is possible, let's go with 100% recovery of Ga from all world ore, with a 10 ppm concentration, to make the math easy. 29,000,000,000 tons, as a world reserve, giving us 29*10000, 290000 tons of Ga available from all the worlds aluminum that we have left, giving us a nice, respectable 846.8 PWh/year maximum possible total world energy production from Ga cells, assuming that the USGS paper is off by a factor of 10 for the Ga input (!), assuming that we mine up every single source of aluminum that humanity knows about (!!!), assuming a near impossible amount of Ga recovered from the ore, and assuming that every single ounce of Ga goes to making solar cells. So I guess if you want to handwave literally everything away, you could make a case. I didn't check the numbers, but thank you for at least doing this calculation. Obviously solar cells aren't going to supply all of the world's electricity and probably gallium arsenide solar panels won't be more common than silicon, so this result doesn't bother me. silence_kit fucked around with this message at 19:31 on Jun 12, 2013 |
|
#
?
Jun 12, 2013 19:28
|
|
|
|
| # ? Apr 25, 2024 20:19 |
|
|
This is probably the weirdest thread I've even seen a slap fight take place
|
|
#
?
Jun 12, 2013 19:34
|
|
