Grid tie to off grid conversion

Ishmael · January 2012

Hello everyone, happy new year. I'm looking for some advice. Here's my situation.

I currently have thirty-eight 185w Suntech panels in a grid-tie setup using Enphase microinverters. The system has worked well for me so far. I generate 23Kwh/day on average, although here in PA the output varies considerably depending on the weather and season. Lately I've been thinking about how I could make use of my panels during a long term grid outage (aka SHTF). I have a Honda EU2000i to handle any short term grid outages.

My goal is to have the proper equipment on hand to manually convert my grid-tie system to a light use off-grid system in response to a serious extended outage. I'd like to be able to run my household size fridge, microwave, toaster, and some electronics (not concurrently) during daylight hours, and just a CFL bulb or two at night. I understand that I'd need the following things:

1.) MPPT style charge controller
2.) Batteries
3.) Inverter
4.) Wire, connectors, breaker/fuses

For the charge controller, I'm looking at the OutBack FM80. It seems like a quality piece that provides a lot of flexibility. Items #2 through #4 are where I'm in need of advice.

Because I'm only planning to convert my system to off-grid in the (unlikely?) event of a long term grid outage, I'm not too keen on spending huge sums of money on lots of batteries that will just sit around degrading or on $2,000 inverter that I may never use. This has me leaning more toward a simpler 12v system. At the moment I'm considering just a single Concorde AGM 108Ah at $300, or maybe even just stealing the 12v batteries from our family cars. Anyhow, this also means a 12v inverter. I'm considering the Xantrex SW2000. It seems okay for $350 and I like that a decent 12v inverter is good to have around anyhow. So for not much money ($600 controller, $350 inverter, possibly $300 battery), I could have the main components on hand to go off-grid during a long grid outage.

The downside of this cheap configuration (aside from the obvious fact that it's small) is that I don't think it will really allow me to leverage the thirty eight 185w panels that I have installed. If I understand correctly, with the Outback FM80 I'll be limited to 1250 watts with a 12v battery system. Does this mean that I'm limited to feeding a maximum of 1250 watts to the FM80 from the panels? Could I wire more (or even all?) of my panels to the FM80 so that I'm practically guaranteed 1250w of constant power delivery even on a cloudy day, since I have all these extra panels to spare? I think this is really the question that I'm most interested in.

My original thought was to wire 3 panels in series (185w x3 @ 105v), and then run two of those in parallel to deliver 1110w @ 105v to the charge controller. Of course, that is on a perfect day - and not even then. Most of the time it would be producing substantially less than 1110w. I supposed that I could manually keep an eye on the weather conditions and connect more panels to the controller only on a cloudy day. But that seems dangerous. What if the sun suddenly pops through the clouds for a minute? Will the charge controller just ignore the excess power delivery? Will it simply trip the 80 amp DC breaker on the battery side of the controller? Do I need a breaker on the PV side of the controller to protect against this? Is this simply a stupid idea?

I have considered the 'big budget' alternative - which would include a couple thousand dollars in a 48v battery bank and an Outback FX3048T 48v inverter, but I'm having trouble justifying the expense for an unlikely scenario. However that would allow me to legitimately feed in 5000 watts from the panels per the Outback FM80 specs.

Lastly, regarding wiring, what is the acceptable gauge wire to run from the 3 panels in series to the charge controller over 75 feet? What about from the charge controller to the battery and inverter? I've seen recommendations for 3/0 or even 4/0 for 12v @ 2000w, but it seems like the Outback charge controller only accepts #2, so I'm confused about that.

mike95490 · January 2012

Re: Grid tie to off grid conversion

Look in to the pricey (but effective) Xantrex XW 600VDC charge controller. Wait, you currently have micro-inverters.

OK, you will need a MC connector Disconnect tool, to un-hook the cables off the micro-inverters from the panels, so you can connect to a charge controller.

As you discovered - a 12V system won't get you far, but you could rig panels into 100V groups, feed your charge controller, and use a small 48V bank of batteries.
At the first sign of SHTF, race to a box store and grab 4, 12V deep cycle batteries. Or the local golf club, and get 8, 6V cart batteries.
You will have to have 48V inverter, Charge controller, MC cables & tools, battery interconnects, fuse, wire boxes, and a room to stash all the gear in, and save the battery purchase for the last moment. Rest of it sits idle.

Or, trash the micro-inverters, and use a hybrid GT system, with battery backup, and you are all set for the next ice storm.

You want 2 plans, you have to pay for 2 plans.

Wire size, look at the caculator spreadsheet that Solar Guppy hosts.

Cariboocoot · January 2012

Re: Grid tie to off grid conversion

Can you say "wiring nightmare"?
Disconnecting the panels from the Enphase inverters and then re-connecting them to feed the FM80 is no small task.
What you should ask yourself first is: "How often does the power go down and for how long?"
Chances are the generator is the most cost-effective power-outage solution.
If you want "quiet power" for outages you could buy a small battery & inverter and use the generator to recharge it during the day (with an appropriate charger which would also keep the batter up when grid power is available).

Can your way be done? Yes, and if you really want to do it will tell you how. But it sure isn't easy.

Ishmael · January 2012

Re: Grid tie to off grid conversion

Yes, I fully expect it to be a big and inconvenient job. I do have a Honda EU2000 generator for any short term outage. This would be only for a long term SHTF scenario - e.g. the grid has been down for several weeks and there is no indication that it's coming back up.

RCinFLA · January 2012

Re: Grid tie to off grid conversion

Based on the stuff you said you want to power you are going to need 3 kWH's to 5 kWH's per day. The refrig alone will be slightly over 1 kWH per day.

The least cost and maintance for long term standby would be a sealed maintance free marine battery. I had really good luck with Delco Voyager Class 27 105 AH batteries. They lasted 7 to 9 years on mostly float charge. Walmart Everstart Maxx Class 30 marine batteries are pretty bad, lucky to get five years.

You can 'Y' connector off microconverter DC panel connections and hook as many in series as charge controller input voltage can take. You leave the series connect down wiring feed breaker open during normal on grid tie feed. For backup you flip off the AC breaker going to micro inverter and flip on the DC breaker going to the series stack PV panels to allow the series stack to feed the charge controller. If you do more then one series string you have to bring down each series string as a separate branch to individual DC breakers, one for each series string. You can flip on all the DC breakers when offline and have multiple series strings in parallel feeding the battery charge controller. Leaving the series 'Y' connections at the panels is fine when operating the micro inverter on grid tie as long as you don't complete the series stack connection path by closing the DC breaker at the breaker panel.

Cariboocoot · January 2012

Re: Grid tie to off grid conversion

So long as you understand that ...

The first step is to determine how much power you will need in an outage. Get a Kill-A-Watt (or equivalent) meter and measure the actual power use of everything you want to keep running when the power goes down. You're looking for two things: what the peak total Wattage is and what the total daily Watt hours are. Remember that things like refrigerators (and other inductive loads, mainly motors) have big start-up surges which have to be taken into account. The K-A-W will not measure these.

Once you have those numbers you'll be able to determine how big of an inverter you need and how much battery to supply it with. That number will tell you how much panel and what charge controller you need to recharge with. You may go back and forth with these numbers a few times, as it is easier to drop "essential" loads than to add power capacity. Cheaper too.

There are other ways around this, including what Mike mentioned; dropping the micros in favour of a central hybrid inverter which has battery capacity built-in. Others will suggest "tricking" the micros into supplying power when the grid is down, but I've yet to hear of sure-fire and painless way of accomplishing this. When power is out you don't want a system that may work ... or may burn up.

Or you could base your system on what power might reasonably be expected from utilizing all of your PV's, which at 7 kW would be considerable. But that means more charge controller capacity and bigger battery bank, which may not be desirable for long-term investment.

Ishmael · January 2012

Re: Grid tie to off grid conversion

Thank you to everyone for these replies. Can you answer this question? Sorry but I had it buried in my lengthy initial post:

>If I understand correctly, with the Outback FM80 I'll be limited to 1250 watts with a 12v battery system. Does this mean that I'm limited to feeding a maximum of 1250 watts to the FM80 from the panels? Could I wire more (or even all?) of my panels to the FM80 so that I'm practically guaranteed 1250w of constant power delivery even on a cloudy day, since I have all these extra panels to spare?

mike95490 · January 2012

Re: Grid tie to off grid conversion

Ishmael wrote: »

Thank you to everyone for these replies. Can you answer this question? Sorry but I had it buried in my lengthy initial post:

>If I understand correctly, with the Outback FM80 I'll be limited to 1250 watts with a 12v battery system. Does this mean that I'm limited to feeding a maximum of 1250 watts to the FM80 from the panels? Could I wire more (or even all?) of my panels to the FM80 so that I'm practically guaranteed 1250w of constant power delivery even on a cloudy day, since I have all these extra panels to spare?

1) you can over supply a MPPT controller, it's supposed to throttle back to save itself

2) My 3,000 watt array, produces maybe 300w on a cloudy day. Clouds = Bad.

Ishmael · January 2012

Re: Grid tie to off grid conversion

Is it hard on the controller when you substantially over supply watts? Or does it simply ignore the excess wattage such that 5,000 watts of excess isn't really any different than 100 watts of excess? My array produces about 6,300 watts on a typical sunny day, for example, so there is potential for a pretty dramatic over supply.

LandKurt · January 2012

Re: Grid tie to off grid conversion

You asked about wire size. If you're still talking about six panels wired as 1110W for 10.5A at 105V I would think that in a SHTF situation you could get by with as little as 14 gauge for a 75 foot run. Sure that would be nearly a 5% loss of power, but in an emergency situation the ability to use a heavy duty extension cord or regular romex household wire could be important. Ideally you'd use properly rated outdoor wiring of probably 8 or 10 gauge.

My question is whether a single 108AH 12V AGM battery could run a microwave or toaster. You'd probably be pulling in excess of 100A with the inverter, possibly 150A for a 1500W toaster. If you only did that while the sun was shining and the controller was contributing up to 80A you might get away with it.

The next question would be can a single Concorde battery deal with up to 80A of charging current. That's way above the commonly recommended 5-13% charge rate. I know AGM batteries can handle higher charge rates than regular flooded batteries, but is .75C doable?

Cariboocoot · January 2012

Re: Grid tie to off grid conversion

LandKurt wrote: »

You asked about wire size. If you're still talking about six panels wired as 1110W for 10.5A at 105V I would think that in a SHTF situation you could get by with as little as 14 gauge for a 75 foot run. Sure that would be nearly a 5% loss of power, but in an emergency situation the ability to use a heavy duty extension cord or regular romex household wire could be important. Ideally you'd use properly rated outdoor wiring of probably 8 or 10 gauge.

This brings up the question of whether this is a permanent install that will be at-the-ready or a plan to implement in the event of grid failure.

My question is whether a single 108AH 12V AGM battery could run a microwave or toaster. You'd probably be pulling in excess of 100A with the inverter, possibly 150A for a 1500W toaster. If you only did that while the sun was shining and the controller was contributing up to 80A you might get away with it.

There's a term for what would happen to a 108 Amp hour 12 Volt AGM if asked to suddenly supply the 100 Amps necessary to run a microwave. It's "electrical collapse".

The next question would be can a single Concorde battery deal with up to 80A of charging current. That's way above the commonly recommended 5-13% charge rate. I know AGM batteries can handle higher charge rates than regular flooded batteries, but is .75C doable?

Concorde claims some unusually high charge capacity. However, all things considered that battery is too small to be at all practical. You would only have 50 Amp hours * 12 Volts or 600 Watt hours max before recharging was necessary. You can go through that much power really fast.

Personally I'd rethink the whole thing, once again based on emergency load needs. That would mean keeping the refrigerator going; 1.2 kW hours per day perhaps, plus a bit extra. I run the cabin on 2.4 kW hours per day. That would be 200 Amp hours @ 12 Volts for a 400 Amp hour battery bank minimum. AGM's are a better choice here as their self-discharge rate is far less than flooded and they can be placed just about anywhere safely.

To that I'll add that the solar should be based on proper recharging. In this case go for the max, since the panels are available. That's 52 Amps peak potential current @ 12 Volts or 624 Watts, an array size of 810 Watts*. You can over-feed MPPT (not PWM) charge controllers some. I would not recommend whopping amounts of too much panel on the idea that a dull day may come. It would be better to configure it so that additional panels could be added in if the particular day was too cloudy. That way you can be sure of not overwhelming the controller.

I would not leave the Enphase inverters connected to the panels when using them to run the battery charge controller; there could be interference with the MPPT function of both the microinverters and the controller in either stage of operation.

*This can be adjusted up or down once the actual efficiency of the site is determined, since there are plenty of panels available.

Ishmael · January 2012

Re: Grid tie to off grid conversion

My thought with the small battery is that I would limit use of high current appliances (microwave, toaster, etc.) to times when the solar is generating full power. Then, assuming the battery was full, I'd be driving most of the heavy load directly from the panels without really crushing the battery.

I spoke to Outback customer support this afternoon. Their customer support person was excellent. As Cariboocoot just mentioned, they did not recommend over supplying the MPPT controller with a huge amounts of watts. If doing so intentionally (such as to compensate for a cloudy day), one suggestion was to place a breaker on the PV side of the controller so that if the sun suddenly popped out unexpectedly, it would blow the breaker instead of the controller. The last thing you'd want to do in a SHTF situation is to blow up your controller trying to eek out a few extra watts.

The responses from this thread and from talking with OB does have me rethinking some things. Now I'm leaning more towards a 48v environment versus a 12v environment, even if it means keeping my eye on the used market for a good 48v inverter. 48v just makes everything easier. I could leverage most of my panels without stressing the controller, I don't need to play around with 0000 wire, and the 48v battery bank would be superior.

LandKurt · January 2012

Re: Grid tie to off grid conversion

It is sure going to be easier to wire up a 48V system to run a microwave than a 12V system. It's nice to work with wire than can actually be bent. Not that I've actually worked with 0000 wire, #2 AWG was tricky enough for me.

But that still leaves the question of what is the minimum battery that could be hooked to a maxed out 80A charge controller. It will be four times as many with a 48V system but the minimum AH would be the same. The 13% maximum charge rule of thumb for FLA would put you at 615 AH, which is a pretty expensive battery system. In an apocalyptic situation four car batteries in series would probably do to buffer power to the inverter while the sun was shining, but how careful would you have to be with charging them at up to 80A?

TnAndy · January 2012

Re: Grid tie to off grid conversion

As you've discovered, 48v is the way to go using the Outback FM80.....I think you can run 4800w of panels, and probably a bit more......I run 1750w ( 10 x175w panels ) @ 24v on Outback 60's, where the factory recommended amount is 1600w. I almost never see over 1600w with 1750w of rated panels.....and typically about 1550 on a good day ( only time it runs over 1600 is a cold, clear day with snow reflection ).

On batteries, since you don't intend to wire this configuration up unless the SHTF, why not buy some good batteries dry, and buy the acid in separate containers to fill them later.....they could sit indefinitely that way. Just make SURE they have never been filled to test, then drained.....because the clock starts running once the plates have had acid on them. Check with local batteries dealers about buying dry batteries.

OR if you can make connections with somebody in the telecom business, do what I did....buy something like the Absolyte AGM batteries.....I got a 1200amp/hr set that was almost new for 400 bucks, the company was going out of business, and my buddy was told "sell everything for scrap price". These are 20yr batteries when just kept on a float charge and not cycled.....which is the case with most of the ones they replace every 4-5 years....most of them are nearly as good as new.

Grid tie to off grid conversion

Comments

Categories