Dumping fully charged Batteries into the Grid

[Seyiwmz]

I'm contemplating an addition to a potential solar installation  this spring.  I'm getting 4kw of solar that's grid-tied, but I'm trying to figure an inexpensive method to add wind to this install without a specialized inverter.  The solar will have a Fronious 5100 watt grid-tied inverter.  I was wondering if i could add batteries that would charge strictly from the  Axial Flux Gen,  and during the night when the sun went down, the inverter could drain the batteries down to 50 percent.  This cycle could repeat everynight as long as there was enough charge in the batteries.  That way the inverter would be in use more than just during the day.  My current windmill overdrives a 3 phase motor now, so i'm not familiar with the inverter type functions, if this is even possible or within there capabilities.  I really would like to build one of the Dan's 17 footers as I've got the book and think it might work better for me.  Maybe that's what everybody else does, but haven't heard anystories of the such..... Hmmm.  Any input would be greatly appreciated, I hate to give up on the wind.

[Phil Timmons]

We were considering something like this, as well. 

Since the PV only gets "use" of the grid-tie inverter a max 8 hours of the day -- but more typically 6 or less, it is just sitting most of the time.

I think you are clever for thinking about adding batteries to "catch" the wind sourced DC while the PV is using the inverter, and then tapping them off through the evening.

We can do all the circuits and designs, and be happy to share them, but I am pretty sure this will completely void all the PV inverter's warranties -- and maybe its UL 1741 listing, as well.

[Madscientist267]

Just a question about this, and I'm not really doubting motives or anything, but...

I'm assuming you would want to dump the battery's 'surplus' to gain points in the grid meter department, therefore lowering your bill.

That part makes sense.

But 50% nightly? Wouldn't the wear on the batteries going down to 50% repetitively outweigh the gains made by the meter?

I would think you're talking about some serious bank here, in more than one way if you catch my drift.

Lifespan = X - (DoD * Cycles) right?

Maybe a lower DoD (down to 70%?) or not as often or combination of both would be more cost effective in the mean time?

Just a thought... Way to think outside the box tho for sure. Props on that. 

Steve

[commanda]

I have to agree with Steve on this one. The cost of wear & tear on the batteries, I suspect, would far outweigh any metering gains.
I'm pretty sure there was some discussion here recently on the cycle life vs depth of discharge for various lead-acid batteries .

For Lithium Iron Phosphate (LiFePO4), which I personally am far more familiar with, it's generally quoted as 2000 cycles at 80% DOD, and 3000 cycles at 70% DOD. Lead acid I seem to recall was orders of magnitude less.

[joestue]

assuming 4000 cycles to 70% dod (i'm using 4000 because one company has bumped up their 3000 cycle claim to 5000, but no one has actually published data yet to prove it!)
i can get 14 cents per kilowatt hour life cycle cost in us dollars.
that's just the batteries not including shipping.

however, 4000 days is over 10 years, do these batteries last that long?

there's no reason to believe that using a second energy source to charge the batteries would void the "solar" inverter's warranty.
worst thing that can happen is you blow the inverter due to over voltage, or you have a ground loop problem fire if the inverter expects to float the solar panels at 60vac relative to the grid ground and you plug in the turbine. and or the turbine shorts to ground accidentally if you were to float the turbine.

[Phil Timmons]

assuming 4000 cycles to 70% dod (i'm using 4000 because one company has bumped up their 3000 cycle claim to 5000, but no one has actually published data yet to prove it!)
i can get 14 cents per kilowatt hour life cycle cost in us dollars.
that's just the batteries not including shipping.

however, 4000 days is over 10 years, do these batteries last that long?

there's no reason to believe that using a second energy source to charge the batteries would void the "solar" inverter's warranty.
worst thing that can happen is you blow the inverter due to over voltage, or you have a ground loop problem fire if the inverter expects to float the solar panels at 60vac relative to the grid ground and you plug in the turbine. and or the turbine shorts to ground accidentally if you were to float the turbine.

I might have misunderstood the concept design (can't really talk without a 1-line and drawing pix, anymore, yunno ), but I was thinking he was talking about using the batteries on the DC inputs (where the PV strings feed-in).  Since the PV is only feeding the inverter 6 to 8 hours a day, the other 16 to 18 hours these string tie points would be fed by the combo of windmill (DC) and stored DC (batteries). 

Since this is not the "intended" use of the Solar Inverter, I am guessing that 10 out 10 Solar Inverter company reps are going to tell you cannot do it.  So of course, us all being what we are we would likely be inclined to go with "don't ask, don't tell."     

As far as the "over-voltage" part, dunno about Frontius, but the inverters we use -- PV-Powered, Solelectria, and SMA-Sunnyboy -- over-voltage on the DC side is a No-Go for the warranty, and the Sunnyboys come-up with a history and tattler of a "Warranty Void" on the LCD screen if they have ever had more than 600 VDC on the string connections.

[Madscientist267]

Well, really it's a two part thing. There's his original intent and question, involving the connections themselves and whether this voids warranty and the like.

I have no idea what the answer is to that. I'd have to go with the 'dont ask dont tell' policy.

But just the very act of asking his question immediately spawned another in my mind: Is it cost effective in terms of battery wear vs meter gains, assuming there are no 'magic smoke events' to invoke warranty proceedings?

and during the night when the sun went down, the inverter could drain the batteries down to 50 percent.  This cycle could repeat everynight as long as there was enough charge in the batteries.

Clearly, he's talking about draining the batteries down on a nightly basis, dumping them into the grid.

Regardless of how this happens (he could be discharging them into normal loads for the 'concern' part), nightly draining to 50% SoC will take it's toll on the batteries.

The argument comes in when the benefits and pitfalls are set side by together and compared.

To me, the pitfalls outweigh the benefits:

1 - The more often, and deeper a battery is cycled, the shorter it's overall lifespan will be. This isn't a big deal on small banks, but on a bank that can even be considered for this kind of thing, there's some serious cash involved every time the bank has to be replaced.

2 - Discharging batteries as a form of "dumping" is also inefficient. There are less losses if the juice is directly dumped to the meter during production, rather than going through conversion, charging, discharging, and more conversion losses to then just be dumped later that night anyway.

3 - If the bank is taken down to 50%, each morning will leave the system with NO reserve, and if the sun doesn't shine/wind doesn't blow, there will be no power to be had (unless one wished to risk further damage to the already stressed battery bank)

It just doesn't make sense. A system should be designed with the exact opposite in mind - keeping the batteries as close to full as much of the time as possible!

Steve
 

[SteveCH]

I add my voice against that much draw-down on the batteries. At least, if they are lead-acid. I don't have the OP's numbers to look at, but he'd have to make a lot of money feeding the grid from the batteries to justify killing them off in a few yr. 20% draw down nightly, that I might think about. But no more than that. The cost of battery replacement would, I'd guess, more than eat up any financial benefit.

Unless we are talking about battery chemistry other than lead-acid. Maybe....

[Madscientist267]

That would be a little better, but I'm thinking only marginally at that. Even with a 20% drain, while the damage is significantly less (that's the maximum I try to take mine down in use), at that point efficiency is even lower.

It's no secret that the last 20% of charge for lead acid is horribly inefficient, so again, better off dumping during production. And discharge isn't all that efficient above 80% either. Lose-Lose IMO.

And if (like many) the batteries never see more than 80% SoC due to the efficiency issues (except for when they see equalization), then 20% off of that puts it down around 60% SoC after the dump.

The more I think about it, the more I find myself discouraging the entire idea. I just can't see any benefits in doing it.

My stand on this is simple: Dump when producing (and full), and save precious battery power (and lifespan) for what it's there for - times when production isn't happening.

Steve

[joestue]

But just the very act of asking his question immediately spawned another in my mind: Is it cost effective in terms of battery wear vs meter gains, assuming there are no 'magic smoke events' to invoke warranty proceedings?

no, absolutly not.
unless you live in california and could sell the power to your neighbor at 20 cents/kwh, then you might break even.

I might have misunderstood the concept design (can't really talk without a 1-line and drawing pix, anymore, yunno  ), but I was thinking he was talking about using the batteries on the DC inputs (where the PV strings feed-in).  Since the PV is only feeding the inverter 6 to 8 hours a day, the other 16 to 18 hours these string tie points would be fed by the combo of windmill (DC) and stored DC (batteries). 

that's what i assumed as well, if the dc is buffered with a battery there should not be any issues.