What would be a good battery to use for renewable energy generation?
December 30, 2009 3:05 PM Subscribe
What would be a good battery to use for renewable energy generation?
If the anual average wind speed of an off-grid house was 6 m/s and had a roof at 30 degrees due south (for the use of solar PV). What kind of battery would you suggest I purchase so that it can be used for a turn key system and last for up to 3 days using wind power and solar PV? - The average power consumption per week of the house is 111.657 KWh
The house is situated in a one hectare (2.5 acre) exposed rural postition.
The house is heated with gas, uses gas for water heating and cooking.
The house is situated in a one hectare (2.5 acre) exposed rural postition.
The house is heated with gas, uses gas for water heating and cooking.
In terms of reliability and price, it's hard to beat lead-acid. In other words, car batteries.
But let's run some numbers. 111.657 KWh / week is to be about 660 watts average (i.e. 660 joules / second).
Your target is 3 days, so it turns out you need to store 172,000,000 joules.
Energy density:
Lithium Ion batteries can store about 700,000 joules per kilogram, and about 900,000 joules per liter. Lead Acid batteries can store about 140,000 joules per kilogram (lead is heavy), and about 360,000 joules per liter
So if you did this with LiIon, it would weigh about 250 kilograms and use 191 liters, which is pretty small. With lead acid, it would weigh 1230 kg (about a ton and a quarter) and occupy 478 liters, which still isn't all that big.
But LiIon can't be used for this application, mainly because the operating temperature range is too narrow for natural conditions, but also because it costs a fortune. Lead Acid works over a much broader range of temperatures and costs a lot less.
Even so, it won't be cheap. Looks like that's maybe 300 car batteries.
Unfortunately, the current (ahem) state of the art in electricity storage is really pitifully bad for high-power applications.
...that's why no one does what you're trying to do. You asked for a "good battery to use". There is no good solution available right now.
posted by Chocolate Pickle at 4:27 PM on December 30, 2009
But let's run some numbers. 111.657 KWh / week is to be about 660 watts average (i.e. 660 joules / second).
Your target is 3 days, so it turns out you need to store 172,000,000 joules.
Energy density:
Lithium Ion batteries can store about 700,000 joules per kilogram, and about 900,000 joules per liter. Lead Acid batteries can store about 140,000 joules per kilogram (lead is heavy), and about 360,000 joules per liter
So if you did this with LiIon, it would weigh about 250 kilograms and use 191 liters, which is pretty small. With lead acid, it would weigh 1230 kg (about a ton and a quarter) and occupy 478 liters, which still isn't all that big.
But LiIon can't be used for this application, mainly because the operating temperature range is too narrow for natural conditions, but also because it costs a fortune. Lead Acid works over a much broader range of temperatures and costs a lot less.
Even so, it won't be cheap. Looks like that's maybe 300 car batteries.
Unfortunately, the current (ahem) state of the art in electricity storage is really pitifully bad for high-power applications.
...that's why no one does what you're trying to do. You asked for a "good battery to use". There is no good solution available right now.
posted by Chocolate Pickle at 4:27 PM on December 30, 2009
Depending on your goals (independence, greenness, money savings?) and the rules in your area, you might consider a grid-intertied system. This gives you a virtual infinite battery for free.
posted by fritley at 4:36 PM on December 30, 2009
posted by fritley at 4:36 PM on December 30, 2009
Oops, you say "off-grid" but then "last for up to 3 days" so I'm not clear on your situation. If my answer is nonsense, sorry.
posted by fritley at 4:37 PM on December 30, 2009
posted by fritley at 4:37 PM on December 30, 2009
It looks like some people are judging the storage numbers on the requirement for 3 days of total drain (ie zero generation from wind or solar during that three day period) which is giving some horrific numbers. This is obviously worst case, based on your description. Is it not unlikely for your application to have zero sun and wind for 3 days, though? I can't imagine solar and wind production would be viable for an area where that scenario was at all likely.
Either your expectations are very high, or your assumed case (3 days at 111kWh total stored energy) is not at all realistic with current battery technology. Perhaps battery storage is not the best solution? Or perhaps you need to factor in your average expected energy production per week (rather than just per year) and crunch that in your better expectations (ie assuming a basic weekly average for continued energy production through the there days)?
I idly looked into alternative energy projects a while back (in no impressive depth more than armchair surfing and TV watching, I'll happily admit) and the best storage seemed to be a stream with a pressure head like mills used to rely on with a water wheel. It seems hard to maintain any decent level of power consumption without very peaky delivery without massive amounts of energy, otherwise. I saw a UK programme where a guy produced most of his power for his house from a mill pond type arrangement like this, and it was very impressive in it's controllability. This may be completely unrealistic, but the numbers that Chocolate pickle come up with seem to suggest a radically different storage style, to me, or at least a rethink of energy consumption.
posted by Brockles at 4:45 PM on December 30, 2009
Either your expectations are very high, or your assumed case (3 days at 111kWh total stored energy) is not at all realistic with current battery technology. Perhaps battery storage is not the best solution? Or perhaps you need to factor in your average expected energy production per week (rather than just per year) and crunch that in your better expectations (ie assuming a basic weekly average for continued energy production through the there days)?
I idly looked into alternative energy projects a while back (in no impressive depth more than armchair surfing and TV watching, I'll happily admit) and the best storage seemed to be a stream with a pressure head like mills used to rely on with a water wheel. It seems hard to maintain any decent level of power consumption without very peaky delivery without massive amounts of energy, otherwise. I saw a UK programme where a guy produced most of his power for his house from a mill pond type arrangement like this, and it was very impressive in it's controllability. This may be completely unrealistic, but the numbers that Chocolate pickle come up with seem to suggest a radically different storage style, to me, or at least a rethink of energy consumption.
posted by Brockles at 4:45 PM on December 30, 2009
I'm pretty sure that lead-acids are still the best deal for that size of application. Not car batteries exactly, but big, deep-cycle lead-acid batteries.
Keep in mind that they do wear out with use, so you should think of battery replacement as an ongoing cost of the system (5-10 years, dunno), with deep cycling being the thing that wears them out fastest. So there's a tradeoff, between deep-cycle batteries (more expensive, but wear out slower), bigger battery bank (means a shallower cycle for a given amount of energy use), and battery replacement. On the upside, lead-acids do recycle well.
links: HomePower, Backwoods Home, Solar Energy International (despite the name they also do micro-hydro and wind stuff).
posted by hattifattener at 4:49 PM on December 30, 2009 [1 favorite]
Keep in mind that they do wear out with use, so you should think of battery replacement as an ongoing cost of the system (5-10 years, dunno), with deep cycling being the thing that wears them out fastest. So there's a tradeoff, between deep-cycle batteries (more expensive, but wear out slower), bigger battery bank (means a shallower cycle for a given amount of energy use), and battery replacement. On the upside, lead-acids do recycle well.
links: HomePower, Backwoods Home, Solar Energy International (despite the name they also do micro-hydro and wind stuff).
posted by hattifattener at 4:49 PM on December 30, 2009 [1 favorite]
my possibly out of date info was 6 v golf cart batteries were the way to go. they are made for deep cycle. Charging way up then fully depleted. Supposedly they hold up best in the kind of usage running a household would be. charging during the day with solar and running at night when you need lights and such. similar to what a golf cart goes through
posted by Redhush at 4:56 PM on December 30, 2009
posted by Redhush at 4:56 PM on December 30, 2009
Keep in mind that it doesn't matter how big your batteries are if your wind turbine and PV panels are incapable of consistently outputting an average of 111.657 KWh per week (and then some for conversion inefficiencies).
posted by JackFlash at 5:19 PM on December 30, 2009
posted by JackFlash at 5:19 PM on December 30, 2009
They'd need to put out a lot more than 660 watts average. You need 660 watts for your usage, plus more generation to charge the batteries. If you figure your power generation at a duty cycle of 35% (which I think is realistic at least some of the time) then you'd need 3+ kilowatts when generation was taking place. (Don't forget that batteries are not 100% efficient. Some of the power used during charging is lost as heat.)
posted by Chocolate Pickle at 6:07 PM on December 30, 2009
posted by Chocolate Pickle at 6:07 PM on December 30, 2009
Consider whether a solar thermal system would significantly reduce your need for electrical output by substituting in your water heating.
posted by biffa at 7:33 PM on December 30, 2009 [1 favorite]
posted by biffa at 7:33 PM on December 30, 2009 [1 favorite]
I've heard the lead acid batteries used for forklifts are the best ones suited for this application.
posted by furtive at 7:34 PM on December 30, 2009
posted by furtive at 7:34 PM on December 30, 2009
hattifattenner's links are good, plus I can recommend The Renewable Energy Handbook, though it's quite North American. If you're in the UK, talk to the fine loonies at CAT for hints.
This is eminently doable, but the load you're describing would be expensive. We'd really need to know the peak load, which I suspect will be at least double your average. Inverters let out the magic smoke if stressed too highly.
There's no way you can get utility power in?
posted by scruss at 8:56 PM on December 30, 2009
This is eminently doable, but the load you're describing would be expensive. We'd really need to know the peak load, which I suspect will be at least double your average. Inverters let out the magic smoke if stressed too highly.
There's no way you can get utility power in?
posted by scruss at 8:56 PM on December 30, 2009
Your question is essentially unanswerable, at least as you pose it, but I'll give it a go…
Some generalities, drawn from my knowledge and experience of power systems in the telco & traction realm:
Beyond the requirements of the load, it also depends on the characteristics of your supply (e.g. peak / average / diurnal / seasonal differences between solar / wind / hydro), the type of controller (& peak/average efficiency of the inverter, if used), and to some degree the type of cells used (yes, the decision-making process becomes a bit circular at this point).
Pure lead, flooded lead-acid cells are pretty much the cheapest (per A/Hr), most reliable, and longest-lasting solution. With predominantly float use (e.g. in a telco situation, or a weekender cabin in a place where they don't freeze / get too hot), you could expect ~10 years life at least, if not 20 years or more.
Forget car/truck/forklift batteries - they may be deep cycle, but the impurities (arsenic etc. remaining from the refining process; antimony etc. added to increase plate strength in vibration-prone vehicle use) reduces the potential A/Hr capacity, maximum discharge rate, and deep-cycle capacity, and dramatically reduces the overall lifespan (particularly, up to a point, with deep-cycle use).
In general, forget most forms of sealed lead-acid cells - their lifespan is poor (regardless of what most manufacturers may claim), their capacity is lower (for a given volume & cost) than flooded "wet" cells, and their deep-cycle ability and maximum discharge rate is considerably poorer.
Consider this: There's a considerable difference in "deep-cycle" usage patterns between starting a car, running a forklift, and supplying a house. In the first case, you're talking about a drain that may last, at worst, a few seconds (high instantaneous discharge rate); you want a battery that can supply up to ~20x its nominal rated capacity (C*20, where C is ~40A/hrs for the average car battery) for a very short period of time without physically (or chemically) degrading too much, and you expect a lifespan of ~3 years of multiple-daily abuse like this.
Forklift batteries are different - lower instantaneous discharge rate (up to maybe C*5 peak when starting; more likely C or C/2 once running), deeper cycle (can be discharged further with less damage), charged at least daily. Again, Most types of sealed lead-acids tend to fit fairly well in this sort of use; sealed "wet" cells (with at least a suitable venting arrangement, and often a catalytic recombiner) are particularly good.
But the requirements of what's essentially a stationary UPS are different again. Here, the arsenic (impurity) & antimony (added for physical strength) found in standard batteries is your enemy, and reduces your overall capacity (at least at higher discharge rates), peak discharge rate, and deep-cycle ability (those last 2 through a variety of chemical and physical effects). Additionally, if you've designed it right, you don't require really deep cycling ability, although pure lead cells are best here too (at least chemically). Your primary power generation (solar/wind/hydro/whatever) should be sized to pretty much supply your daily requirements, plus overhead to charge your batteries (at up to maybe the C/2 or C/5 rate, depending on various things).
So what you require (in a lead-acid cell, at least) is high purity lead (no arsenic), no added crap (antimony) for mechanical strength, and sized (along with the primary power system) to the size and type of load. Stationary traction engine start batteries, 'solar' lead-acid batteries, and telco-style flooded cells, are the easiest types to get a hold of.
Back of the envelope calculations for your numbers (assuming 12v-120v inverter) suggest a ~600A/hr (@ C/10 rate, the usual rate for measuring lead-acid capacity - note, not the car-battery-style CCA rating) @ 12v system, and you'll want a primary supply that can generate a daily average of maybe 10A~12A @ 12v. From that, and Ohm's law, you can extrapolate what might roughly be required for a 6v/24v/36v/48v system.
But, as I say, that's very roughly…
posted by Pinback at 9:41 PM on December 30, 2009 [1 favorite]
Some generalities, drawn from my knowledge and experience of power systems in the telco & traction realm:
Beyond the requirements of the load, it also depends on the characteristics of your supply (e.g. peak / average / diurnal / seasonal differences between solar / wind / hydro), the type of controller (& peak/average efficiency of the inverter, if used), and to some degree the type of cells used (yes, the decision-making process becomes a bit circular at this point).
Pure lead, flooded lead-acid cells are pretty much the cheapest (per A/Hr), most reliable, and longest-lasting solution. With predominantly float use (e.g. in a telco situation, or a weekender cabin in a place where they don't freeze / get too hot), you could expect ~10 years life at least, if not 20 years or more.
Forget car/truck/forklift batteries - they may be deep cycle, but the impurities (arsenic etc. remaining from the refining process; antimony etc. added to increase plate strength in vibration-prone vehicle use) reduces the potential A/Hr capacity, maximum discharge rate, and deep-cycle capacity, and dramatically reduces the overall lifespan (particularly, up to a point, with deep-cycle use).
In general, forget most forms of sealed lead-acid cells - their lifespan is poor (regardless of what most manufacturers may claim), their capacity is lower (for a given volume & cost) than flooded "wet" cells, and their deep-cycle ability and maximum discharge rate is considerably poorer.
Consider this: There's a considerable difference in "deep-cycle" usage patterns between starting a car, running a forklift, and supplying a house. In the first case, you're talking about a drain that may last, at worst, a few seconds (high instantaneous discharge rate); you want a battery that can supply up to ~20x its nominal rated capacity (C*20, where C is ~40A/hrs for the average car battery) for a very short period of time without physically (or chemically) degrading too much, and you expect a lifespan of ~3 years of multiple-daily abuse like this.
Forklift batteries are different - lower instantaneous discharge rate (up to maybe C*5 peak when starting; more likely C or C/2 once running), deeper cycle (can be discharged further with less damage), charged at least daily. Again, Most types of sealed lead-acids tend to fit fairly well in this sort of use; sealed "wet" cells (with at least a suitable venting arrangement, and often a catalytic recombiner) are particularly good.
But the requirements of what's essentially a stationary UPS are different again. Here, the arsenic (impurity) & antimony (added for physical strength) found in standard batteries is your enemy, and reduces your overall capacity (at least at higher discharge rates), peak discharge rate, and deep-cycle ability (those last 2 through a variety of chemical and physical effects). Additionally, if you've designed it right, you don't require really deep cycling ability, although pure lead cells are best here too (at least chemically). Your primary power generation (solar/wind/hydro/whatever) should be sized to pretty much supply your daily requirements, plus overhead to charge your batteries (at up to maybe the C/2 or C/5 rate, depending on various things).
So what you require (in a lead-acid cell, at least) is high purity lead (no arsenic), no added crap (antimony) for mechanical strength, and sized (along with the primary power system) to the size and type of load. Stationary traction engine start batteries, 'solar' lead-acid batteries, and telco-style flooded cells, are the easiest types to get a hold of.
Back of the envelope calculations for your numbers (assuming 12v-120v inverter) suggest a ~600A/hr (@ C/10 rate, the usual rate for measuring lead-acid capacity - note, not the car-battery-style CCA rating) @ 12v system, and you'll want a primary supply that can generate a daily average of maybe 10A~12A @ 12v. From that, and Ohm's law, you can extrapolate what might roughly be required for a 6v/24v/36v/48v system.
But, as I say, that's very roughly…
posted by Pinback at 9:41 PM on December 30, 2009 [1 favorite]
Ooops...
"… a primary supply that can generate a daily average of maybe 10A~12A @ 12v in addition to the ~60A @ 12v your household load requires.
posted by Pinback at 9:47 PM on December 30, 2009
"… a primary supply that can generate a daily average of maybe 10A~12A @ 12v in addition to the ~60A @ 12v your household load requires.
posted by Pinback at 9:47 PM on December 30, 2009
(Not so good with numbers today…)
"… suggest a ~6000A/hr …"
posted by Pinback at 9:52 PM on December 30, 2009
"… suggest a ~6000A/hr …"
posted by Pinback at 9:52 PM on December 30, 2009
You know what would be small, cheap, and reliable? A gasoline-powered generator. If your average use is 660 watts, a reasonable generator could produce that on a couple of gallons of gas per day.
That's why most people looking for cheap, reliable backup power for moderate power levels (like yours) use gas or diesel generators.
Disney World has backup power, in case Florida's power goes out (which has been known to happen during hurricanes). Their power plant can provide enough power for the entire park for several days. But what they have is gas turbines, drawing out of huge tanks of natural gas. That's because they need hundreds of megawatts to keep everything going.
One of the huge advantages of petroleum and natural gas is that they're cheap, plentiful, readily available, and the energy density is excellent. Battery technology just isn't there yet for this kind of thing. A lot of very motivated chemists are working on the problem, but it's non-trivial.
posted by Chocolate Pickle at 10:13 PM on December 30, 2009
That's why most people looking for cheap, reliable backup power for moderate power levels (like yours) use gas or diesel generators.
Disney World has backup power, in case Florida's power goes out (which has been known to happen during hurricanes). Their power plant can provide enough power for the entire park for several days. But what they have is gas turbines, drawing out of huge tanks of natural gas. That's because they need hundreds of megawatts to keep everything going.
One of the huge advantages of petroleum and natural gas is that they're cheap, plentiful, readily available, and the energy density is excellent. Battery technology just isn't there yet for this kind of thing. A lot of very motivated chemists are working on the problem, but it's non-trivial.
posted by Chocolate Pickle at 10:13 PM on December 30, 2009
Response by poster: Sorry for such a slow reply, I live in the UK, so my day time varies from your's a fair amount.
The system has to be offgrid as it would cost too much to fix it to the grid.
It would be best if the battery could last for three days in case of days where there was little or no sun/no wind e.g. on an overcast day with little or no wind.
posted by sockpim at 2:44 AM on December 31, 2009
The system has to be offgrid as it would cost too much to fix it to the grid.
It would be best if the battery could last for three days in case of days where there was little or no sun/no wind e.g. on an overcast day with little or no wind.
posted by sockpim at 2:44 AM on December 31, 2009
Why isn't your turn-key system suggesting/supplying a battery? This would almost be like a turn-key intranet computer that didn't come with a hard drive.
posted by DU at 3:36 AM on December 31, 2009
posted by DU at 3:36 AM on December 31, 2009
what about cracking hydrogen with excess power, and using a fuelcell to turn it back to electricity. On low power days, you could use your natural gas hookup for electricity generation as well.
some examples:
http://www.hydrogencarsnow.com/blog2/index.php/hydrogen-economy/itm-power-homes-could-be-carbon-negative/
http://www.renewableenergyinternational.com/release/renewable-hydrogen-home-permitted-be-constructed-grand-cayman
posted by jrishel at 7:16 AM on December 31, 2009
some examples:
http://www.hydrogencarsnow.com/blog2/index.php/hydrogen-economy/itm-power-homes-could-be-carbon-negative/
http://www.renewableenergyinternational.com/release/renewable-hydrogen-home-permitted-be-constructed-grand-cayman
posted by jrishel at 7:16 AM on December 31, 2009
Jrishel, storing the hydrogen is problematic. The energy content of hydrogen burned in a fuel cell is 1.62 megajoules/kilogram, so our hypothetical 172 megajoules requires about 106 kilograms.
At one atmosphere, hydrogen weighs about 0.09 grams per liter. So storing that much hydrogen at one atmosphere would require a storage space of 1.18 million liters. Put a different way, it's 90 grams per cubic meter, and storage would require just shy of 1200 cubic meters, which is a cube about 35 meters on a side.
Compressing it isn't really a good answer, either; the compression process uses a lot of energy, and it's all loss.
Hydrogen as a fuel has quite good energy per mass, but it's really dreadful when measured per volume.
posted by Chocolate Pickle at 2:40 PM on December 31, 2009
At one atmosphere, hydrogen weighs about 0.09 grams per liter. So storing that much hydrogen at one atmosphere would require a storage space of 1.18 million liters. Put a different way, it's 90 grams per cubic meter, and storage would require just shy of 1200 cubic meters, which is a cube about 35 meters on a side.
Compressing it isn't really a good answer, either; the compression process uses a lot of energy, and it's all loss.
Hydrogen as a fuel has quite good energy per mass, but it's really dreadful when measured per volume.
posted by Chocolate Pickle at 2:40 PM on December 31, 2009
By the way, if you do store that much hydrogen at one atmosphere, keeping it from floating away isn't trivial. And it's a fire hazard. (You've just built the Hindenburg.)
posted by Chocolate Pickle at 2:43 PM on December 31, 2009
posted by Chocolate Pickle at 2:43 PM on December 31, 2009
(You've just built the Hindenburg.)
Only if you've painted your storage tank in rocket fuel grey.
posted by notyou at 4:05 PM on December 31, 2009
Only if you've painted your storage tank in rocket fuel grey.
posted by notyou at 4:05 PM on December 31, 2009
Ah, rats. No, it's a cube about 11 meters on a side. (I did a square root, not a cube root.)
Still kind of a problem to build.
posted by Chocolate Pickle at 7:12 PM on December 31, 2009
Still kind of a problem to build.
posted by Chocolate Pickle at 7:12 PM on December 31, 2009
Chocolate Pickle, it's a house not a car. Burying a large propane-type tank or two to store the hydrogen isn't that crazy of an idea, and you'd need some amount of pressurization just to move the hydrogen through pipes. Hydrogen is the same type of a fire hazard then natural gas or propane and people keep tanks of that stuff around with hardly any problems.
posted by jrishel at 6:14 AM on January 6, 2010
posted by jrishel at 6:14 AM on January 6, 2010
sockpim, I can think of very few places in the UK that an offgrid solution would ever be cheaper than running a line in - unless you're on a wee island.
posted by scruss at 2:51 PM on January 9, 2010
posted by scruss at 2:51 PM on January 9, 2010
This thread is closed to new comments.
None of this is going to help you in the immediate future, but something to keep an eye on for a replacement system I guess.
posted by tad at 3:25 PM on December 30, 2009