diy solar

diy solar

My first experience running off batteries during power outage:

thr0ttl3j0ck3y

New Member
Joined
Apr 10, 2020
Messages
23
We recently spent a day without power here in NW Montana. I have been building up a battery backup out of cheap 6v flooded lead acid batteries I've been buying from NAPA as cores. Initially, I was worried about how well they would work, as they've supposedly been turned in because they're no longer any good. But after reconditioning them, and running them to power my 20' camper trailer for a day, (I live in it full time, year round) I now realize I was worried for nothing. I only have 6 6v deep cycle batteries, and they powered me through without a hitch. I didn't even have any solar hooked up to them yet, I'm still in the process of salvaging the wiring I'll need to hook up charge controllers. I ran them as a 2s, 3p 12v battery bank. I know, there should be some pretty big losses in the system with 12v, but my trailer is already set up for 12v, so it was just a no brainer to set it up that way and skip the expensive inverter to run a higher voltage battery bank and power my trailer through the shore plug. I love it. I'm also planning to try hooking the whole set of batteries in series to try charging the 36 volt bank directly off one of the 44v nominal panels. I can build a current limiting device with a tupperware, some salt water, and two chunks of metal, and test the voltage output and move the plates in the tupperware until I have a safe charging voltage. I'm not gonna just leave it and trust it, I'll be checking it as it charges, but during the power outage, I figured LOTS of things out I had been dreading, and this stuff isn't quite as finicky as I had thought. My voltage didn't even get below 12.2v after a full day of running off the cheapo used batteries with no charge coming from panels. Once I get the panels installed, I'll be golden. Anyhow, just wanted to share my experience, and maybe give people some reassurance that you don't have to have all the best expensive batteries, or crazy control circuitry, if you have a decent understanding of some of the basics, you can do a lot with very little money.
 
Thanks for sharing your experience! It's helpful to us newbies to see real working examples, and I'm sure you feel rewarded for your engineering and it's subsequent success. Keep plugging along!
 
FLA batteries are stone age (rock solid?) technology and are fairly dummy proof. Awesome for learning. I have a used golf cart with six used 8V Trojan batteries in series. And a 48V UPSverter that I back-feed my breaker panel during grid down.

Monitoring the power while accelerating or climbing hills can easily top 100A. Likely your battery cores were slow for golf cart service, but have lots of capacity left for minimal camper usage. I plan to switch to Nissan Leaf when the FLA are too slow for the golf cart, but will keep as a secondary battery bank for extended grid-down. We lost power for 1.5 weeks during hurricane Ike.

Look for a MPPT solar charge controller that can input 44v nominal and output 12V to the batteries. You can then power the camper while charging. I’m sure the board here can make recommendations.
 
Last edited:
@thr0ttl3j0ck3y, thank you for sharing your experiences. Just a couple of things to consider:

I know, there should be some pretty big losses in the system with 12v, but my trailer is already set up for 12v, so it was just a no brainer to set it up that way
You have no need to increase the voltage of your battery system - 12V is absolutely fine. My rule of thumb when considering the battery system voltage is:
  • 1w to 3,000W load: 12V
  • 3,000W to 5,000W load: 24V
  • >5000W load: 48V
I'm also planning to try hooking the whole set of batteries in series to try charging the 36 volt bank directly off one of the 44v nominal panels.
You do not need to increase battery system voltage in order to use high(er) voltage PV panels when using an MPPT SCC (Maximum Point Power Tracking Solar Charge Controller). I would not recommend using a PWM-type SCC these days - they are old tech and should be consigned to history. Your PV array could be producing hundreds of volts and will still safely and efficiently charge a 12V battery.
 
@thr0ttl3j0ck3y, thank you for sharing your experiences. Just a couple of things to consider:


You have no need to increase the voltage of your battery system - 12V is absolutely fine. My rule of thumb when considering the battery system voltage is:
  • 1w to 3,000W load: 12V
  • 3,000W to 5,000W load: 24V
  • >5000W load: 48V

You do not need to increase battery system voltage in order to use high(er) voltage PV panels when using an MPPT SCC (Maximum Point Power Tracking Solar Charge Controller). I would not recommend using a PWM-type SCC these days - they are old tech and should be consigned to history. Your PV array could be producing hundreds of volts and will still safely and efficiently charge a 12V battery.


I have plans to get one eventually, but I'm doing all this with basically zero money. Bit by bit, I'm getting it done, though.
 
At this very moment I'm enjoying my first power outage with my shiny new Sunny Island 2S2P system.

During the night I heard thunder and saw diffuse flashes through the clouds. Heard a bit of rain. Just now heard a nice close "Crack!"
Woke up at my usual 6:00 AM, went for coffee an saw the power light was dark, although machine was still warm. Difference between battery and AC clocks indicated an hour without power.

Flipped the interlocked breakers in main panel and power restored by batteries. Just double-checked, and yes I remembered to turn off the breaker feeding "grid" to Sunny Island. Otherwise, it would see grid restored and after 5 minutes monitoring switch back to grid, then anti-islanding or just plain grid failure would cause it to go off-grid again.

Sunny Boy 10000TLUS-12 was doing its job with minimal sun, but SWR2500U showed grid failure due to frequency. It doesn't have the right parameters to support frequency-shift power control, and Sunny Island has default settings where it spends time at 59 Hz to keep mechanical clocks at correct time, on average. (Needs to be configured to not drop below 59.5 Hz if standard older GT inverters are to be supported.)

At 9:00 AM the remaining inverters and their panels, oriented toward 2:00 PM and 5:00 PM sun, are producing 1 kW or 8% of peak power, due to oblique sun angle and partial cloud cover. Battery was supplying 900W, realized espresso machine was still on so I turned it off. Present loads are four full size refrigerators and one tube amp, plus laptop and monitor. Should be just about zero charge or discharge of battery at the moment. Once PV capability reaches 3 kW I'll turn A/C on and make things comfortable on this muggy day.

Two hours after PG&E tweeted they had enough power to avoid rolling blackouts, they started the first rolling blackouts since the Gray Davis era. Those aren't so bad, 1 hour per community. But some non-rolling blackouts have also occurred. Oakland's waste treatment plant was caught by surprise and dumped 50,000 gallons of raw sewage in the bay. Guess the did the math and decided, "Fines are cheaper than Generators." Other waste treatment plants run on multiple generators, some using methane from their own digester tanks. Due to where they are located and water flow, they aren't allowed any dumps of untreated sewage. Oh, I guess Oakland was just doing what they could get away with and not have a penalty?

Anyway, it's nice to know my tens of $k investment has a return in feeling smug and satisfied. :)

update 9:45 AM: Turned on A/C. Sunny Boys delivering 4 kW (SWR2500U still offline). Sunny Island putting 300W into battery. Can't read label on A/C due to overgrowth, but estimate rest of house consuming 1 kW and A/C consuming 2700W. PV production probably curtailed by frequency shift because no more power needed at this time.

Prior outages that only affected my street (power to the underground transformer in my front yard feeds 6 customers) weren't known to PG&E's system, but this time the map shows status of customer outage reports all over the Bay Area:


1597595210050.png
 
Last edited:
I guess this is a "Mixed Feeling" situation.

I don't think I ever want a power failure, but if we spend all this money for battery backup, we want to see it do the job at least once, right?

I am justifying the cost on being able to time shift power and save money even with the grid up. With the close fire here, we have had the lights flicker quite often, but it has not cut out yet. My battery is sitting and waiting for the inverter to arrive, and then I have to wire it all up. It is a pretty sure bet that the power is going to fail before I get the inverter wired up. The Lake fire is still just 12% contained and burned almost 18,000 acres.
 
I am justifying the cost on being able to time shift power and save money even with the grid up.

I'm thinking in terms of protecting the food, both commercial and home-harvested, in the freezers and fridge. A/C is sure nice too!
Once I get the full UPS configuration wired up, it will also give me protection from brownouts, which have claimed some of my electronics. (A simple relay and power-good circuit would also accomplish that.)

Have you done the math? My math says batteries cost more than buying power from PG&E, regardless of time of use rates.

Power costs me $0.15, $0.30, $0.45 /kWh depending on time of day.
AGM batteries cost $0.40/kWh to replace after reaching the manufacturer's spec'd cycle life.
FLA costs $0.20/kWh (and demands more of my time.)
PV costs me $0.05/kWh.

So I think the most cost-effective is to produce excess power, backfeed the grid when it is up and curtail production when grid down (just bounce along fully charged batteries.) Then turn off all but essential loads at night during grid failures.
 
I guess this is a "Mixed Feeling" situation.

I don't think I ever want a power failure, but if we spend all this money for battery backup, we want to see it do the job at least once, right?

Power was restored. :(

I guess I just enjoyed the most expensive cappuccino in the world.
I suppose whatever they serve on the ISS costs more, with lifting things into space priced at $500k/pound, but technically that's out of this world.

My machine, however, at $1300, has paid for itself many times over. Figure I'm down to $0.12/cup for capital cost.
And it's always waiting for me to brew the moment I get up.

1597606339194.png
 
Have you done the math? My math says batteries cost more than buying power from PG&E, regardless of time of use rates.

Power costs me $0.15, $0.30, $0.45 /kWh depending on time of day.
AGM batteries cost $0.40/kWh to replace after reaching the manufacturer's spec'd cycle life.
FLA costs $0.20/kWh (and demands more of my time.)
PV costs me $0.05/kWh.

So I think the most cost-effective is to produce excess power, backfeed the grid when it is up

I am currently back feeding the grid around solar noon, but then buying it back at a much higher rate in the evening. For me, this is as much about not getting ripped off on the "Time of Use rate" than it is trying to truly profit. I think SCE makes plenty of money, and does not need to take more from me. They take my extra power and profit from it, and then want me to pay for it also? But I will also try running the numbers again.

With new batteries, I agree, they really can't save money over the cost of grid power. If I use the cost of my inverter and batteries (and a rough estimate of all the little stuff) I am in about $6,000 thanks to a great deal from Battery Hookup. My pack measures about 17 KwH of useable storage. They are basically new cells and I should get a solid 3,000 cycles of 60% DoD. So 10 KwH x 3,000 = 30,600 KwH for $6,000 = $0.20 per KWH cycled. That is right in the ballpark. If I move my over production to the peak rate $0.38 time, it can actually start saving some money.

I truly think the utilities should be paying home battery owners more when we are actually helping to stabilize the grid. We become the peaker plant to shove power out to the grid when it is needed most. In the past, that was the most expensive power for them to generate as it needed to go online fast, and if it ran at a loss, it was accepted to keep the grid up until they could ramp up more coal etc. If people with solar and a battery can hold off noon production by storing it, and then provide the boost needed in the evenings, we should get paid a premium, not billed extra. If I was able to stay on my old tiered billing, I would be able to stay in tier 1 90% of the year. But no, have to be "Time of Use" now, and the cheapest rate is more than tier 1 was. So we use less power, they have to charge us more for it. Thanks, NOT!

Our water company did the same crap. They demanded we all use less water, handed out penalties for using lawn sprinklers, required all new home fixtures and appliances to use less water, and then, they complained that revenue was down, so they increased the fixed charges and the cost per gallon so we ended up paying the same amount (or more) for less water used. How great is that?

Eventually my battery bank is going to starts dying. At that point I only have to replace the batteries, not the inverter and the rest of the install. My hope is that I can get another battery bank then (in 5 to 8 years by my math) that will match that cost per KwH cycle. Maybe the tech or used EV cells will get cheaper by then. That was another reason for the Schneider inverter. If some new battery comes out, or I go to a different chemistry, I should be able to program it to work just fine. Up at 48 volts, it should be easy to find a grouping of cells that fall in the operating range of 44 to 60 volts. Even my BMS can take cells from 2.0 to 5.0 volts each, and up to 24 cells. So I can try all kinds of cells.

The other fun thing I am looking at is the new bidirectional EV chargers. Right now you can get a unit that will work on the Nissan Leaf. It can charge the car at up to 7,000 watts, but it can also pull from the car at 7,000 watts to feed your home. This turns your car into a 60 KwH power wall. Now if you are at work with the car while the sun is shining, it does not help too much. But for my brother, he and his wife both have Chevy Bolts, hers is home and could charge on the sun, and then when he get's home, push half the free solar charge into his. Too bad they don't have a bidirectional for the Bolt yet, it uses a different connector and the data link is not supported yet. My current car is only a hybrid, not a plug in yet, but it is 7 years old with 175,000 miles, so I am looking. Being able to use the car battery to power my home will be on my list of options.
 
I agree on the power rates, and the water.

Water company said, "We don't want to lose the momentum" i.e. we enjoy the boosted rates that were necessary to pay for operations during the drought, and now want to charge the same selling prices when water is plentiful.

As for power, PG&E and PUC say that the last kWh purchased during middle of the day is what our power production is worth. Actually, for end of year surplus they pay us $0.04, probably average of all power from around the clock (neglecting our reduced consumption and surplus production.)
I want to call a solar strike for all PV owners. Yesterday at Noon would have been perfect. Simultaneously take all our capacity off line.
Let PG&E buy power to make up the shortfall, and the price they pay for the last kWh is what each of our kWh are worth.

As for batteries, looks like Lithium can be cheaper than AGM. Not planning to cycle, buying fewer cycles was the way I chose to go. $5000 for 20 kWh of SunXtender, usable capacity 14 kWh at 70% DoD, 650 cycles. That's $0.55/kWh, lasts about 20 months. More total Wh for shorter cycles, 10 year float life. Using 15% DoD for 3500 cycles might balance it for 10 year life, and would represent a battery bank sized for 3 days autonomy. But it would still cost $0.48/kWh and take 10 years to recoup.

Your 3000 cycles is 8 years, so time-value of money probably means at best you broke even.

That's why my plan is give PG&E a kW when it is worthless to them, for a credit of $0.15 but cost to me of $0.05. Actually give them 3 kW, $0.45 credit and cost to me of $0.15. Then later in the day, take one kWh back from them and return the $0.45. I use PG&E as a 33% efficient battery at zero capital cost, rental price $0.33/day. My power is almost free, costs me $0.05 for off-peak and $0.15 for on peak.
 
That's why my plan is give PG&E a kW when it is worthless to them, for a credit of $0.15 but cost to me of $0.05. Actually give them 3 kW, $0.45 credit and cost to me of $0.15. Then later in the day, take one kWh back from them and return the $0.45. I use PG&E as a 33% efficient battery at zero capital cost, rental price $0.33/day. My power is almost free, costs me $0.05 for off-peak and $0.15 for on peak.
This kind of thinking makes me seriously wonder if someone in the bowels of PG&E's ugly grey HQ is laughing at you, and at everyone with grid-tied solar. With the current regulatory climate, you will never get a decent price for resold solar power, even at peak usage times. Many people have installed grid-tied solar thinking they were gonna make a nice profit selling juice to the utility---and been disappointed at the results. Read the official CPUC document on home solar, if you haven't seen it already. Lots of warnings about unreliable solar installers but very little about resale pricing.

 
That was an interesting read. It is like they really don't want you to put in solar.

NO! You must buy our power. If you make any power, it will NEVER pay off.

My solar was up and running for about 3 weeks before my NEM 2.0 actually took effect. That first month was by far my cheapest as the system still charged me the old tiered billing. I was not trying to be sneaky, the city inspector turned the system on after inspecting all the wiring, and he took current measurements etc. I expected him to shut it back off, but he said, no, just leave it up, it's fine, SCE has the paperwork. Any delay is on them. My first bill after the NEM agreement went into effect was just $6.00 but that is only the taxes and "non bypassable charges". It charged my about $80 onto the new full year power plan. Since I sized my system to only match my needs, I didn't expect to ever have a negative month, so I wanted to stay on monthly bills. I was never asked, they just put me on the full year billing. And since it has been $40 to $90 each month, that year total is going to end up about $600 I have to pay in one lump. I would much rather pay $50 each month, and it is changing back to that, but they could not change it until the full year ended. Only a week left now. Now with the battery going online soon, I may end up with a few negative months, but I doubt it, they are so good at finding things to charge us for. We are still paying extra for them to shut down a nuclear power plant???

After my system was planned and ordered, they (SCE) have changed the ToU rates twice, each time making it worse for solar. Had I known where it was going, I would have probably put up a few more panels to cover the gap. The reason I didn't was that 16 panels maxed out my breaker panel with the 20 amp back feed breaker. At the time, I thought I would have had to change out the whole panel, and that was a very big expense. I didn't realize then I could have stepped down the main breaker to a 90 amp and then back feed up to 30 amps.

With all my essential loads moving to the panel after the XW inverter, I just may have to do that to be legal. I will have 40 amps of circuits which will be supplied by solar and the XW batteries, but when there is no sun and the batteries run down, it will need to power those circuits from my back feed breaker in the main panel. Making that a 40 amp feed to a sub panel is totally safe and legal. But since it is also the solar back feed, it brings up the buss bar current limit. The solar will still have only a 20 amp breaker into the sub panel. The XW will be set to never export more than 10 amps or so to the grid (to zero out loads in the main panel, not out to SCE) and never when the solar is exporting. So I am not worried about it being safe. Just trying to stay legal. Of course, the electrical codes are written for the "worst case" to still be safe. If there was a programming error, the solar panels could be pushing their 16 amps, and the XW could crank out 28 amps, and all the essential loads all happen to be off, then there could be up to 44 amps going back to the main panel. I am running #8 wire but I am going to leave it on the 20 amp breaker for now. If something like this happens, it will obviously trip the breaker. This is why I may have to bump it up to a 30 amp and drop the main to a 90. And that will also help to ensure my essential loads can feed off of it. The XW will be set to provide current when the loads pull more than about 10 amps over the solar production. So unless it runs the batteries down, it will actually be fine even on the 20 amp breaker. The XW has very good features for helping a weak grid feed or a small generator. I will just have to take some measurements for a few days (weeks) and make sure I have a decent balance and can keep enough charge in the batteries. I am hoping I do not have to split the circuits and can just move 4 existing breakers to the essential loads, but if it comes down to it, I can climb back in the attic and split off things that are not "essential".
 
My rule of thumb when considering the battery system voltage is:
  • 1w to 3,000W load: 12V
  • 3,000W to 5,000W load: 24V
  • >5000W load: 48V
I have much more conservative take on the same chart. I just dont want to have to size wiring and lugs for 250a of current. ~100a is were I draw the line. so the same chart for my personal take is
1-1200w -12v
1200-3000 -24v
3K-up - 48v
 
1-1200w -12v
I personally bump this up to 3,000W in my rule of thumb because of the faff, additional cost and additional wiring required for the 24V-to-12V converter that would be required for any 12V loads. Without considering this, I too would probably settle on 1,200W-1,500W being the 'natural' switch-over point.
 
I personally bump this up to 3,000W in my rule of thumb because of the faff,

A valid point.

If you want to carry the high current torch, Ill carry the low current - DC-DC converter one and we can recommend people evaluate both for their needs ;)
 
Someone went on a rampage in a nearby street a few weeks ago smashing a car into power poles. They managed to break one, knock another at an angle and ended up on their side against another. Power was off for 12 hours from about 3am. I didn't even notice the mains was off, solar installation kept everything running without any break.
 
Someone went on a rampage in a nearby street a few weeks ago smashing a car into power poles. They managed to break one, knock another at an angle and ended up on their side against another. Power was off for 12 hours from about 3am. I didn't even notice the mains was off, solar installation kept everything running without any break.
Good chance you were the only one on the street who had lights. And very few of your neighbors had anything, not even a generator. My attempts to talk neighbors and friends into buying at least a SMALL solar/battery power box, like a Jackery or Bluetti, or even a cheap generator, have been mostly useless. "I can't afford it" is the usual excuse. They have money every month to pay PG&E tribute but they can't manage $125 for a generator. Wonder if most human beings are simply not smart or determined enough to create their own electric power.

The fire situation around Nor Cal is insane right now. Rolling outages are being threatened.
What does the Bay Area media tell people? "Get your candles and hand fans ready, a return to the Dark Ages is imminent."
 
Last edited:
I did hear a portable gen running later that day, probably keeping someone's fridge and freezer running. Several people in the neighbourhood have them to carry them should the mains fail as a cyclone sweeps through/past. You are right though, my house was the only one in the area that had lights in small hours. There are plenty of solar installations but they'd all be grid tied with no battery.
 
Back
Top