I can't imagine spending the money on solar and having no battery at all, since I consider lighting to be greatest utility for the least watts of all possible applications of electricity!
Mark, as you describe, there are workarounds for many things. As Joseph and Rebecca have pointed out, flashlights/headlamps go a long ways, and you don't necessarily need a lot of watts for decent lighting. I recharge my 8650s(lithium-ion flashlight batteries) in a USb-powered charger that can run off a car adapter, or a laptop, or any 12V system with a cig-lighter plug for my USB adapter. It is truly amazing how far a flashlight or two worth of watts can go, in a room with a nice reflective white ceiling.
However, I am fond of the ability to run power tools after dark, especially when 'dark' is 4PM for part of the winter here... It's a good time for this sort of work since the garden is put to sleep, and it may be unpleasant outside... enough power to work in the evenings is a big help IMO.
Couple things to consider about minimizing or eliminating the battery bank...
1) A properly maintained battery's lifespan in cycles will be vary based on depth of discharge... so a 100AH battery discharged daily down to 50% will wear out notably faster than a 200AH discharged daily to 75%, and then on rare occasions discharged to 50%. Of
course, since batteries also degrade with time, there's bound to be a sweet-spot...
2) Rate of discharge impacts how much you can draw from a battery; a '100AH' Trojan 27-AGM is rated thusly:
5-Hr Rate 77AH
10-Hr Rate 82AH
20-Hr Rate 89AH
100-Hr Rate 99AH
So the higher the draw relative to the bank capacity, the smaller the percentage of AH you can actually draw...
3) Solar charge controllers are really designed to charge batteries, and don't necessarily operate correctly without them in the loop. The cheapo on my folding panel won't work at all without a battery; it will provide 12V power to 12V devices connected to it, it NEEDS a battery connected as the controller determines what to put out based on the battery voltage. I think this is fairly common behavior...
If you're going to skip the charge controller, then you better make sure whatever you connect to your panels has a broad input range that can handle the the varied input it will receive... And even then you may get some interesting behavior when clouds pass over the sun and you suddenly have a power deficit with nowhere to draw from...
John Weiland wrote:The latter problem could be somewhat ameliorated by having them placed in some sort of earth-bermed shelter that would buffer the temperature swings more.
Yup, this would be a really good thing for efficiency. I found one example that observed the consumption of a chest freezer decline from 3.04KWh/day to 2.14KWh/day on lowering the ambient temp from 99f to 75f... and a room kept at roughly earth ambient would be even cooler than that that. Plus, the room would be a great place to store lots of food that doesn't need refrigeration but would benefit from the cool temps.
John Weiland wrote:
The transfer switch at our power pole allows me to disconnect the mains from the utility and use my generator for times when the power is out. Assuming enough solar panels (+/- complementing wind turbine) were installed and were producing enough AC power (via inverter) to feed into the house, one might simply disconnect from the utility in the morning and use the panels by day to run low-load appliances and re-power a few batteries. High-load appliances would be turned off during this period. Does anyone know of any rural cooperative utility law (USA) that would prohibit one from doing this on a routine basis? Does anyone's rural coop have a clause(s) about minimum KWh usage or frequent disconnect that I'm not aware of? Thoughts about an automatic transfer switch that might do this?....Clearly, this would be different than a "true" grid intertie set-up.
I might hit this problem from another angle; practicality would really depend on the house/property/usage though. Maybe it would be possible to move lighting and low-draw stuff over to the solar/wind system, with separate wiring? Then use a transfer switch(much smaller and cheaper than the one on the pole since it wouldn't need nearly as much capacity) at the house to switch this stuff back to grid in the evening if/when needed. The high load appliances run on grid whenever desired...
John Weiland wrote:Just a FWIW, it's probably been mentioned before, but there are those small thermoelectric "six-pack" coolers that draw ~50 - 60W that may be of use for some small refrigerator-storage applications.
The efficiency of a peltier-based cooler is absolutely *terrible* compared to any phase change refrigeration system I've ever heard of, even the ammonia-filled absorption cycle sort that RVs tend to get since it can run directly on propane. IIRC they're about 1/4 as efficient as a compressor based fridge. Plus they will not be able to hold a useful temperature in a hot environment.
There are DC powered compact compressor based fridges and freezers available, including compact/portable units used in boats and 4x4s(ie National Luna, Arb, Engel, at the low-end Waeco), and these are generally very efficient, but not cheap. There's also the chest-fridge option of converting a chest freezer to operate as a fridge, which is quite efficient due to the orientation and better insulation.
John Weiland wrote:
I rigged up a string of 150W incandescent flood lamps (8 bulbs) over a remote space for feeding some geese and drove these with a 12V/~30Ah battery running through a 1500W modified sine wave inverter.....would leave the lights on for several hours while they fed and was pretty impressed with the output, duration, and robustness of the system.
I think I'm misunderstanding something here! 8x150W bulbs? AKA 1200 watts... for several hours? A 30Ah 12V battery has 360 Watt-hours in it till flat dead... So something is interesting about this setup!
Joseph Lofthouse wrote:Capacitors are an alternative power storage mechanism. They could easily hold enough electricity for LED lighting into the evening, and can power surge-loads during the day.
I think some beefy capacitors would probably be a very good idea for a system trying to economize on batteries, and would certainly help with startup surges especially for larger loads. However, my understanding is not so practical for storage... let's see...
Per engineering toolbox, energy stored = 1/2 C V^2
where C=capacitance in Farads, V=voltage
So, a cheapo (16V rated, run at nominal 12V) 83F ultracapacitor bank off ebay for $115 would give: 0.5*83F*13.8*13.82=7903 Joules
So, that's
2.2 watt-hours for $115. Enough to run a decent LED for an hour...
One problem is that this is to discharge to 0V; the amount of power actually available would be less, because at some point the voltage will be too low to be useful. If you're direct-driving LEDs, maybe this is 6V? In practice, you'd probably have a LED driver of some sort to manage the brightness. This is probably going to want at least 2.5V, though I do see one rather expensive option good down to 1V. So, some power will be lost from the voltage dropping to an unusable level, however you set things up.
(There's a list of drivers here:
http://www.videofoundry.co.nz/ianman/laboratory/research/driverlist.php)
However, the real problem is a single 18650 lithium-ion rechargable battery is good for at least
8 watt-hours, and will cost less than $15 each even for high amperage capable protected cells...
(I don't much like the idea of series capacitor arrays, especially not the cheap and chinese sort; I would at a minimum greatly underrate such caps before using them... but it's just an example. Quality caps would make it even worse value.)