Is there any reason why capacitive storage could not be incorporated into the solar panel itself? I am thinking of domestic solar applications. When solar energy input exceeds household demand the excess could then go into storage. When this is full the excess could be switched to the grid. The household could use the capacitive storage preferentially to drawing from the grid.
A system designed to produce approximately the same energy as the household uses on a 24 hour cycle would contribute minimally to grid demand fluctuations and perhaps approach independence from the grid.
Let’s look at roof solar panels: for example these
Each one is 240 Watts. Let’s say 12 hours a day of sun, and you want in-panel storage capacity of 30% of daily output - sounds reasonable? That’s 864 Wh that you would like to store. The best gravimetric energy density for capacitive energy storage (for Ultracapacitors) is 5Wh/kg. So you will add 172 kg to the panel’s weight. That means you will increase each panel’s weight 800% when adding capacitive battery to it. Volumetric energy density is about 6 for Ultracapacitors, so you will be adding 144L of volume to the panel - increasing the panel volume by about 300%. And the additional cost is probably going to be huge, although the table in the link above has some weird info for it.
tl;dr - no way can you incorporate modern capacitive energy storage do what you want. Maybe after it is improved by a couple of orders of magnitude.
Because generally the energy storage density (J/kg or J/L) or capacitors is of order 1/10 that of even rechargeable batteries, which itself is low compared to, e.g. liquid fuels.
Thanks Terr and Carl Pham. That’s just the answer I was after.
I didn’t know what the energy density of capacitors is. Like so many ideas in this field, it seems like a neat and tidy solution and then slams headlong into practical realities.
J.
Great answers from the Physics and Science perspective.
Here’s the engineering perspective :
Design a system to store the excess energy that will **give an return better than the bank rates 
**
So lets do the math here :
Selling 1kwh to the grid will get you 7c (and this is an optimistic number). Since power is produced only 12 hrs in a day - you can at the max 1x7x12 = 84c/day per kwh. Lets assume as a best case that all days in a year are sunny. So max you can earn in a year is 84c x 365 = $307 per kwh
Lets assume a 5 year return time (most good investments yield returns in the 2-3 year range)
So $343x5 = $1535
So a system that stores such that power is delivered to the grid at the switch of a button, continuously 365 days a year (365 cycles at least for a battery) and needs no investment on maintenance, has a life more than 10 years (you want to make a profit -right ?) and costs less than $1535 per kwh delivered back to the grid (you get paid not on the power stored but the power put back on the grid) will make you a little profit in the sixth year of investment :).
I have a 5kW system on my roof. I get paid 8c (AU) for each kWh fed into the grid. However I pay something like 27c for each kWh I draw from the grid and this figure is going skyward very quickly. (91% increase over the past 5 years.)
My system is economically sensible not because of the 8c I get but because I can draw power from it for my own use during the day which saves me pulling from the grid. It also insures me against future price rises.
Now if I had some storage solutions then I would be able to use a greater proportion of what I make and be nearly self-reliant: only drawing power on extended cloudy periods. (And in the heat of summer when I make 32kWh and use only 14, I would still be able to sell the excess.) My use of a battery would help the power companies too since I would not be contributing to daily supply fluctuations.
As good sense as it is, the price of battery tech does not make the investment feasible.
For solar-thermal (i.e. mirror arrays driving a steam turbine), excess can be stored in insulated tanks of molten salt - although some heat is inevitably lost that way, there are no thermodynamic losses from conversion to another form of energy - the molten salt heat storage is apparently 93% efficient.
That is of course not the efficiency of the plant itself as a whole - obviously the efficiency of any heat engine is limited by Carnot’s Theorem (as well as the practical, even lower limits imposed by engineering etc)
For a low tech solution I think that lifting a heavy weight and then releasing it has some merit. Old cars would be impractical, but surely a 100 tonne block of concrete (A 4 metre cube?) would be possible, together with some lifting mechanism, either hydraulic or mechanical (block and tackle).
The engine would provide the power for lifting and run as a generator when required.
Is that likely to be more efficient than just pumping water uphill?
You’ve fallen into the trap many homeowners do. A “good investment” that returns the initial investment in a two year time interval is equivalent to a bond which pays 50% interest and has near-zero risk. That would be a truly awesome investment. A 3 year return equals 33% per annum, and 5 year return equals 20%. (I’m ignoring compounding and inflation for simplicity). Mean time the “bank rate” now, even for 10 year bonds, is about 2% and net of inflation may turn out to be negative.
Bottom line: when we demand our physical engineering financially outperform our craziest most fraudulent financial engineering by a factor of 3x or more, well we’ve set up a much-too-high barrier to innovation and improvement.
A more realistic ROI for a long-term project is inflation plus 5%, i.e. something in the 7-10% nominal range. Sticking with zero inflation for simplicity as you and I did above, that means a 5% ROI equals a 20 year time to payback (TTPB) is a plausible financial target.
Admittedly at some point you need to be concerned about the lifespan of the equipment; a 20 year TTPB doesn’t work if the device itself fails & needs to be replaced after 15.
Economists’ surveys have discovered that many people will demand ridiculously high ROI on unsexy long term purchases which really ought to be thought of more like financial investments. In other words, your flawed analysis is in good company.
We now see smart financial companies starting to exploit this interest rate arbitrage opportunity between bank rates and the actual ROI of solar. The business borrows money at 2%, buys & installs solar panels that “pay” 10-15%, and gives the homeowner a smidgen of the margin for their trouble.
Finally, as j_sum1 points out in his second post, for the particular economic situation of personal power generation vs. inelastic personal demand, there’s a further benefit in the difference between the cost avoided in not buying some power at retail prices vs. being able to sell any excess back at only wholesale prices. An economic analysis which ignores that cost avoidance is ignoring a hefty chunk of the benefit available.
Well, the concrete block won’t evaporate in the desert air but the flip side is the probability that hydro generation is more efficient that the gearing necessary to produce electricity from a slow dropping weight.
And just to throw another technology into the mix - Compressed Air Energy Storage
If you happen to have an appropriately sized salt dome underneath your solar plant you can use that as a pressure vessel.
You could go a step further and release the compressed air pressure through a gas turbine and make the turbine much more efficient via the added inlet boost.
In fact, solar power itself is limited by Carnot’s theorem, although in that case the Carnot limit is high enough that some other limitation almost always becomes much more relevant. A solar panel is, after all, extracting energy from the temperature difference between the Sun and the Earth. It’s a form of heat engine, even though it may not look like it.
Could you combine the two? Use the excess power to pump water from a storage tank into a tank with a floating lid with a 100-ton weight on top of it. When you need power, open the check valve and let the stone push the water through a turbine to generate power. It has the advantage of being closed loop except for normal parasitic losses. It’s also not dependent on terrain (although it would require a fair amount of space).
I have no idea where you are getting that figure from, the report clearly shows (p. 6) that you will get about 1 GWh/year per 4 acres or so. For ten acres that works out to be 2.5 GWh or 2,500 MWH/year, or about 7 MWh per day, so off by a factor of 10.
I am sorry, is the title of the page I gave you “It Takes 2.8 Acres of Land to Generate 1GWh of Solar Energy Per Year, Says NREL” or am I mistaken?
Yes, but it links to the report I posted and the 2.8 acre number is for a centralized plant PV 2-axis station of greater than 20 MW. To do that you would need, minimally, at27% efficiency on the plant, about 90 acres.
From that link: “A large fixed tilt photovoltaic plant that generates 1 GWh per year requires, on average, 2.8 acres for the solar panels. This means that a solar power plant that provides all of the electricity for 1,000 homes would require 32 acres of land.”
Is 32 acres less than 90?
Well, there is some semantics here. There is what they call “Direct area” which is the area needed for the panels themselves, and then there is total land use which includes area between the panels, access roads and such, converter stations etc. So lets go with the 2.8 acres per GWh/yr. Now the average home uses about 10 MWH/yr so for a 1,000 homes you need 10GWh. So 28 acres. That’s cool, so how much capacity is that? Well at about 27% efficiency that is about 5 MW. But that 2.8 acre number is for plants 20 MW or bigger so minimally it seems you would need 112 acres. Since I have used approximations for all my calculations it is in the neighborhood. The headline isn’t wrong, but it needs context.
What if the lid of the water storage tank was a float and you dribbled in water from the top that (eventually) floated the weight up to the roof, and then attached the weight to a cable, drained the water (making power) and then lowered the weight (making power). Would the power from the weight be “free” so to speak.
I have a feeling it would take a lot of power to move the weight.