Theoretically

[fabricator]

So, lets say for the fun of it a guy had 44 280 watt panels and the wire run from the array to the house was 750 feet (to get away from shading) what are the options for the wire run?

[dbcollen]

How long is a string?

You need to provide all the pertinant info if you want an answer. You have been on this board long enough to know that.
info needed:
panel VOC, VMP, ISC
controller quantity, max volts, max amps

[XeonPony]

well at 2Kv you can use 14 guage! that will gaive ya a 1.65% Voltage drop

At 500V  6 guage will give ya 3.69% volt drop

100V  800KcMill  will give ya 3.05%   (That is basicaly a small copper tree sizes wire !!!)

Best bet use a good inverter system and plumb it highest voltage you can get to your house!

[fabricator]

I know next to nothing about solar, so I have no idea about how many panels in a string, these are the specs for the panels.
280W
280
44.2V VOC
36.6V VMP
8.26A ISC
7.66A IMP
77” x 39” x 1.78"

[fabricator]

well at 2Kv you can use 14 guage! that will gaive ya a 1.65% Voltage drop

At 500V  6 guage will give ya 3.69% volt drop

100V  800KcMill  will give ya 3.05%   (That is basicaly a small copper tree sizes wire !!!)

Best bet use a good inverter system and plumb it highest voltage you can get to your house!

The system would probably use a Xantrex 4048 with Classic controllers.

[XeonPony]

Well at 750 feet it doesn't matter as it becomes a game of voltages and at that power idealy you'd want it  to go Array > Inverter > 120(240) primary to 600V secondary > Cable run > 600V primary to 120(240) secondary > House loads.

Around here 240 to 600v transformers can be had with a 48 pack off beer and a local line man damned things get lost all the time!

End game is as big voltage as you can safely get to the house for keeping cable size reasonable.

[fabricator]

Ok, I'm starting to see it now, there is a local electrical surplus outfit that has literally tons of transformers they sell very cheap.

[oztules]

On the wind vs solar thread, I make mention of the grid tie/home grid.

If it turns out that it is as advertised, then you may be able to run the panels voltage back to the house at 600vdc, and have the inverters (grid tie and home unit and batteries there).

If it does not do as advertised, then yep, inverter at the panels and bring home the AC at elevated voltages as Xenon suggests.

Hopefully I will know more and what brand/s types can do this. I'm fairly certain at this time that it was the German ones that could be programmed for this, but it may be more common than I know at the moment.

If they can get rid of a lot of other components and wire gauge, they may even be seen as cheap.... who knows.



..............................oztules

[fabricator]

sounds interesting alright.

[ChrisOlson]

The most panels you could run in series would be five with a Classic 250.  That would put you at 220 volts Voc with 30 volts of headroom for cold weather to HyperVOC.

5 in series gives 183 volts Vmp.  A 12 kW array would have 65 amps under ideal conditions.  That can be done, but it will take big wire.  My guess is that you'll spend $5 Grand just on the wire run if you run it at voltages that a Classic 250 can handle.  Stepping it up to 240 and transmitting it with grid tie inverters at the panels won't make it much better.  You're still at 50 amps.

I think you need to be at least 600 volts to get the amps down to 20 or so in order to afford the wire.  At 600 volts you won't be buying Classics.  You'll need Xantrex XW-MPPT80s:
http://www.solar-electric.com/xaxwmp80amp6.html

It also means you'll have 500 amps going into your bank on a perfect day (you said XW4024, which is a battery based inverter).  To handle 500 amps you'd need 7 of those big Xantrex controllers.  That's $8 Grand just for controllers.  At 500 amps charge rate on a 24 volt system you need a battery bank sized at 5,000 amp-hours.  My best guess at cost on a 5,000 ah 24V bank would be about $20,000.

If you intend to grid tie this you need three 4024 inverters.

It's a really good thing it says "theoretically" in the subject line.
--
Chris

[tecker]

I'd stay around 200 to 250 also and use the wind turbine home run to combine the power . I have a 100 foot run with super caps at 60 ft in Pvc box and the voltage drop is negligible running 24 volt .

[oztules]

It would help considerably if you signal your intended use for this....Yes, your turn for not enough info

If you want to transport the full 12000 watts down the line, then  grid tie, you could buy a HV inverter and run 600dc direct to the inverter.

If you want to transport the full 12000 watts down the line to charge a monster battery bank...., or use 12000 watts (good lord why), then your in the pain territory Chris has outlined. Invert at the panels, batts at the panels, and transformer transmit to your chosen load and transform down again... expensive every which way.

If you want to run a house and charge a 700ah battery bank.... use 48volt system, not 24v......now we'll find out why......
For this application, you will likely need to transport only 5kw-6kw max I would suggest.... full charge and load of 2kw should do it.
The best equipment for this will be the SMA range of inverter chargers, and SMA grid tie inverters.

I can now confirm that the fellow who came here today, uses the hv dc to the grid tie inverters. They directly parallel the output from the SMA 240vac 48v inverter driving the house..... yes it is that simple I'm afraid. .... He lives 500 miles away, and monitors it all on his phone with 30 channels of logging...... all facets of the system.... so I saw it in real time from way over here.

The grid tie units will power your house AND charge the batteries through the house inverter at their proper settings for the bank size etc.  including equalise if needed. No other mppt controllers or gadgets required.

Sounds too good to be true... soooooo........ I couldn't help it...

I got my no-name inverter 48v 70amp charger 18000 watt surge, and decided to put my money where his mouth was.
I hooked the grid tie units I have (I bought a pallet of them for peanuts 1.5kw CEHE units) straight onto the output of the inverter, and then stood back and waited for the grid tie units to log on and start....... not sweating much I might add.....much.

I figured I could rebuild the output stage of the big inverter if it all turned to mush, and I have plenty of grid tie inverters (I really bought them for the S/Steel box they are in... ).. so if anyone should try this at home, I guessed I was it.

The output stage circuitry for this big inverter is available on the net, so a rebuild should be straight forward. The charger power and battery type is programmable for 8 different types of technologies, and then equalise if wanted.... and it is up to 70 amps (you get to set A  max).... Thats why we want 48v. The current limit is for either 24 or 48v versions.... so 48v wins easily, as can charge at twice the power as the 24v versions.

Well I turned on some loads so the grid ties would have something to work into, and all went as expected, and as Chris has seen too.

But  because of the H bridge output design, as I turned off the loads, and watched the amp gauge on the battery to inverter rail... I saw it slowly move from discharging the batteries, to charging the batteries... even with 500 watts of load still on line.

I took it up to 20 amps battery charging before I decided enough was enough for 1 day..... and I was running out of solar, darn clouds came back over, but I had 1.5kw from the panels into the load AND the batteries........ Hmmmm looks promising.

Dont  try this at home unless your prepared to kill an inverter (I doubt it, but I don't want folks to blow up their stuff until we know more about this.)

All worked as advertised, and should work for all inverter chargers using big output transformers and H bridge outputs ( should be all).

Also, I prefer the heavy transformer galvanically isolated grid tie units.... more flexibility  and because I do.

So The next and only problem is where does the power go when the batteries are full, or the load drops off and the batteries are near float etc etc.

We know the grid mppt controls the power from the panels by running up the va curve until they are happy with their Pmax... so that stops excess power coming from the panels.... but below this level

For the SMA range, it is simple and seamless, no work required apparently. It simply shifts the Hz a fraction, the grid ties see it, and runs up the VA curve  further to limit the current.. that simple and Auto.

For the rest of us:
Simply turn on a 240vac pwm load of up to 4kw (hot water then radiators) to act as a dump load till we figure something better. Maybe drop off a grid tie if the battery hits float voltage...  Need to monitor the system set up like this to find what parameters will give us the best switch points to isolate the grid ties in sequence if no dump loads are used.

We also need to research the SMA inverter chargers to see just how they do as I am told they do.... but the system works as is, but as the kw of the panel goes up, the SMA looks the best set and forget system out there at this stage.

I don't think Xantrex has this facility... so is a non starter.



..................oztules

[oztules]

Ok, the SMA sunnyisland 5058 6048-us would fill that position. It will control everything provided you use sunnyboy grid tie inverters.
It has to be coded by the distributor for this purpose, but will seamlessly run as we want to see it in this theoretical instance. Totally control battery charging and input requirements. We need care less about voltage drop in the line, as it will be a who cares if it dont heat up thing.... we can afford plenty of losses provided they are not dangerous. (heat)

It will not ever draw more than the 5kw (or 6kw) from even 100 kw of panels.. (but would start at sun up with 5kw till sunset if required) This will be the way of the future for solar off grid now that solar is soo cheap. We can do now what we only dreamed of 3 years ago.

Max DC is 550vdc for the 5000 series sunny boy.

Not cheap, but very well backed up, and unbelievably good system.... xantrex may or may not offer this.... I doubt it, but you can hope

I suspect xantrex and others will need a dump load system, as the only viable way to control the power in the system.

We either shed it in the panels, or waste it at the end... take your pick.

I'll have to take the latter, as my grid tie units are only $70.00 each 1,5kw,(UL listed CEC TUV etc)  and the inverter charger was only about $760.

SMA is the best for non fiddlers, and plug and play.
 I would suggest this course not mine unless you like to play up a bit.

the manual for the 5048 to get your head around this.
http://www.energymatters.com.au/images/SMA/SI5048-11-EE3107_Prosa.pdf


....................oztules
Edit: If the sunny island sees the batteries charged, it increases frequency by 1hz..... no action... 1.5hz, sunnyboy decreases power by 50% 2hz, sunny boy turns off and linearly everywhere  in between +1 and +2 hz.... problem solved. (see manual above )

or:

For a bang bang system(turn off turn on non-sunnyboy/island (mine) SMA suggests turn off at battery charged voltage, and let bank discharge 5% and then turn on again... repeat as necessary.

or:

Or just use a dump load (hot water etc)

[ChrisOlson]

I suspect xantrex and others will need a dump load system, as the only viable way to control the power in the system.

My SW Plus shut down with an error when I tried to backfeed it with a grid tie inverter.  As long as there was plenty of loads to keep the grid tie setup busy, everything was fine.  But the grid tie thought the big Xantrex was the "grid", and it must increase voltage or frequency or something that the Xantrex didn't like.
--
Chris

[fabricator]

OZ, my intended use is pretty much exactly what yours is, to continue living at my present level of use and just throw away power if I need to, grid tie with battery back up, that system sounds to good to be true.

[ChrisOlson]

Fab, grid-tie with battery backup setups don't run all the power thru the battery bank.  You use use those little grid tie inverters that mount on the back of the solar array and they all feed into a Net Metering setup with a separate meter.  So you bury a service to the array and the utility comes and installs a meter to measure your power you feed back to the grid.

At the house end you have your battery bank and the utility power feeds the AC1 input on the battery-based inverter, and the generator feeds the AC2 input.  The batteries are charged off the utility power, not by the solar directly.  If the power goes out the solar shuts down, the inverter automatically switches to battery backup to keep your power on, and the generator auto-starts if the bank gets too low before the power comes back on.

You CAN hook up a wind turbine or another solar array to charge the batteries and "sell" to the grid with the battery-based inverter too.  But that's just on very small systems.

There is no exotic "mini-grid".  The utility will require that your Net Metering system automatically shut down in the event of a power outage, and they will require you to install (at your expense) a lockable disconnect switch at your Net Metering point so they can manually shut it down and lock it out at their discretion.

That's how every big grid-tied setup I've seen is done.
--
Chris

[ChrisOlson]

I'll have to see if I can dig out some of the pictures I've got of the big FIT 20 kW solar installations I've seen in Canada.  There's a big array on a tracker and there's a little (I think they're EnPhase) grid-tie inverter for every 4 panels mounted right on the back of the array.  All those inverters feed into a combiner.

Hydro (what they call the power companies in Canada) buried a service to the array and there's a stand there with Net Meter on it.  That feeds all the power back to the grid and gets your "credit".

The house is a 1/4 mile away and is a standard grid home.  The power "credit" from the array is used to offset the usage at the home's meter.  Installing the battery backup system inverter (like Xantrex XW) and battery bank is optional, and only used in the event of a power outage to keep the lights on.  But in most cases the battery inverter doesn't pay.  You can just install a standby generator instead to keep your lights on in the event of an outage.
--
Chris

[fabricator]

Ok, that makes sense, then you have no dump loads or clippers or any of that stuff for direct grid tie. And since I make biodiesel at $0.80 cents a gallon I just use my 12kW diesel genset when the grid goes down.

[oztules]

There may be some confusion here Fab.
The system I described is fully self contained. Any referenceto the sunny boys or grid tie inverters is for home grid ... not connected to any power grid.

The "grid" is simply the sunnyisland with built in inverter charger. If we hook the sunnyboy grid tie to any amount of power (solar ), it will self regulate to charge the batteries through the sunnyisland unit, run the loads if there are any present, and idle down to just supply thepower to the sunnyisland if there are no loads.

It is seamless. The sunnyboy has the mppt, and converts DC to AC sensed from the sunnyisland. The sunny island drives loads on the AC out if the sunnyboy has not enough power, or redirects it to the batteries via internal charger if there is an excess. If there is too much power from the sunnyboy into the sunnyisland, then the sunny island throttles the sunny boy back until they are in correct power supply/drain/charge. It all happens auto.

It can also use the real grid the same way. The grid powers the loads, the sunnyboy outputs it's full power into the real grid, the sunny island fills the batteries.

So it is flexable, and pretty bullet proof.

Any or all problems Chris suffered with his experiments to not have any bearing on this sunny island/sunny boy combination.

I don't know anything about the xantrex units Chris.... but if they incorporate internal inverter chargers, then they should charge the batteries if you have a grid tie inverter driving the output (AC).

If it failed as soon as grid tie input was greater than the loads, and not charge the batteries, I suspect it is in the programming inputs, not the inherent design of the H bridge output.
Xantrx should be able to put you in focus with what it can really do with reverse feed situations..... poor unit if it cant. For that money I would hope it could keep up with the german product.... or even the Chinese product I am testing out now, both put excess power from the output AC terminals into the battery.... nifty really.

The ten minute test worked for me yesterday, and I was planning on doing more extensive tests today driving 70 amps into the batteries from the grid tie inverter....... but the darn sun has disappeared again... and so has the wind so don't say it .. probably only doing 800 watts from solar at the moment, pretty dull, and not enough to excercise the inverter seriously at all. Enough to easily power the house, but not to charge the batteries seriously at the same time.... where are my new panels!


.................oztules

[fabricator]

That sounds really good, I've never liked the grid tie thing, might just be paranoid but I'm sure you get screwed somehow, I think I'll let you do some more smoke tests though. but to have something like that that's totally set and forget would be like hitting the lotto.

[DamonHD]

I have long been interested in the SMA SunnyIsland / BackupBlocker stuff, and I hope it would play nicely with my 5kWp of PV on 4 SunnyBoy inverters.  Our grid is good here, but if it's off I would definitely want to be charging batteries when the sun is out.

My primary use would either be to shift loads to lowest grid carbon intensity (kgCO2/kWh) and/or to ensure that we import from the grid at all when daily generation drops below daily consumption (Nov--Mar typically) and export all the rest of the time.

But I'd need Li batteries for the round-trip efficiency to be worth doing on CO2 grounds, and I'd need SMA to actually talk to me rather than ignoring me completely as has been the case so far on this and other matters.

Rgds

Damon

[ChrisOlson]

I don't know anything about the xantrex units Chris.... but if they incorporate internal inverter chargers, then they should charge the batteries if you have a grid tie inverter driving the output (AC).

I don't think they can backfeed from the AC output to charge batteries.  They'll only charge from the AC1 (grid or generator) input, or the AC2 (generator only) input.

We have dual generators on ours.  The one on AC1 has a much tighter tolerance for input freq and voltage and cannot be adjusted for out-of-spec.  The one on AC2 can be adjusted to a pretty wide voltage range, but freq has be from 55 to 63 Hz or it will spit the generator off.

They're synchronous inverter/chargers but I don't know if that means they can charge batteries by having power forced into their AC output.  I don't think so because our inverters (SW Plus) do NOT have grid tie capability so they can't "sell" to the grid (assuming a person has a grid to hook to).

boB would know.  He designed those inverters.
--
Chris

[oztules]

"They're synchronous inverter/chargers but I don't know if that means they can charge batteries by having power forced into their AC output.  I don't think so because our inverters (SW Plus) do NOT have grid tie capability so they can't "sell" to the grid (assuming a person has a grid to hook to).

boB would know.  He designed those inverters."

My chinese thing  has no grid tie provision either, but the thing easily charges the batteries from being force fed from the rear.... so to speak. (AC out).

The Sunnyisland takes full control and maintains a very strict battery charging regime for normal inputs, I THINK it does the same thing when back fed, .
But I think it is quasi control.. ie I don't think/know if  it controls the battery charging in reverse feed,  or is it just using the inherent design of the H bridge.... the same way regenerative braking works  in motor controls, and controlling the sunny boy instead to suit the requirements it thinks the battery needs... I don't know, that is just conjecture from me.

Hopefully boB can fill in the myriad of missing gaps.

For mine, it is a high power FET H bridge synthesising the 50 hz waveform in the incredibly lossy output transformers (with filters etc).
In charger mode (grid tie backfeeding through the transformers) it simply seems to work as a controlled rectifier controlled by pulse width... in other words power flows both ways depending what we do with it.

( one of the big differences with xantrex units and mine will be decent transformers in the xantrex, and big but high magnetising currents in mine. I will rewind some very very big torroids for my unit and bring the idle current down from a whopping 200w to maybe 25 watts.... just by changing the output transformers ( and a filter or two. .. The unit it tough, cheap and lossy.... but is easy to work on.... easy when you say it quick.

For me it is going to be suck it and see experimentation, but I fully expect that reverse feeding in my chinese inverter will just blindly charge the batteries if the load does not need it, (not sure what the cpu is doing at this time) and I will have to intervene by taking control of the buck converter in the mppt of the grid tie units, to overide the mppt signal and drive the VA curve up the scale when I need it to back off.... or just trigger a turn off pulse, and shut it down till the batteries use up 5%, and turn it on again. Thats what SMA suggest for their range of grid tie units (older ones) that can't sense what a change of frequency means.... my first try will probably do this as it's inherently simple... I like that.

The fact that they change frequency means the output of the sunny island can never be connected to a real grid, being used in this way, as it would conflict with the grid frequency.... and we all know who wins that contest.



Edit:

 DONT DONT DONT try this with inverter chargers that are HF (High Frequency /light weight) inverters. I have no experience with them in this application, and because of the HF PWM switching of the tranny  I  don't even remotly expect them to back feed to a battery bank.... blow up... yes... feedback ..no.

 
Low Frequency Transformer coupled inverters  and galvanically isolated grid tie inverters are VERY VERY heavy, mine weigh over 110 lbs, Chris's is probably in the same region. If your inverter weighs less than 70 lbs plus it is probably NOT a suitable candidate for this process.




...................oztules

[ChrisOlson]

Low Frequency Transformer coupled inverters  and galvanically isolated grid tie inverters are VERY VERY heavy, mine weigh over 110 lbs, Chris's is probably in the same region. If your inverter weighs less than 70 lbs plus it is probably NOT a suitable candidate for this process.

Our new SW Plus 5548 that I just put in on Tuesday weighs about 145 lbs, and the Outback auto transformer above it that supplies the 180 degree out-of-phase leg of our split phase power weighs about 50 lbs



I'm pretty sure our particular model can't be back fed into it's AC input.  It just sees a problem with that and shuts down with an error for whatever reason.  It has generator load support in it so when either generator is supplying power to it on AC1 or AC2 inputs the inverter inside it somehow synchronizes with the generator's frequency and voltage.  The generator power is pass-thru to loads and the remainder of it goes to battery charging.

However, let's say the generator input is 30 amps.  The house loads are 10 amps.  The charger in it will use 20 amps to charge the battery.  Now, if we turn on something big, Real Big, in the house that draws 100 amps surge the inverter will instantly switch directions.  All the generator power goes to the load (40 amps on surge with our generator) and the inverter contributes 60 amps of it from the batteries.  And it does this so fast the lights don't even blink in the house when just a split second earlier it was using 20 amps from the generator to charge the batteries.

The way I understand it, this is called a "synchronous bi-directional inverter/charger".  There are only a few of these ever built.  Xantrex SW and XW's can do this.  Outback's new GS8048 Radian can do it.  The SMA sunnyisland can do it.  And Magnum Energy will be coming out with a model that can do it sometime next year (so I've been told by some people at Magnum) and there is a reference to it on the internet:
http://www.nacleanenergy.com/articles/13695/hybrid-inverter-charger

Outback FX-series inverters cannot do this.  Neither can any of Magnum's current line of MS-series.

So maybe from that description, oz, you can figure out what's in mine because I don't have a single clue how inverters work.  I only know it shuts down if incoming power from the turbine exceeded loads and the turbine tried to backfeed it as it would the "grid".  I don't remember what the error code was, but I think it was for frequency and voltage out of limit.

Edit:
oz, the other thing is that might be important is that these SW Plus inverters have three huge transformers in there.  They're on the end of the inverter where the control panel is.  That end of the inverter weighs about 100 lbs and the other end is a lot lighter.

I have no clue why they got three transformers in there.  Most other inverters I've seen, like Outbacks, only got one.  So I don't know if they got the extra ones because of dual AC inputs or what.
--
Chris

[oztules]

Yep, mine has two monsters in it (low grade methinks) and are identical.

The primaries are run in series, and the secondaries are run in parallel. I fully expect yours are the same but three instead of two... basically 3 x 2kw/transformers all paralleled for max power handling on the output. The three outputs in parallel lower the impedance and spread the heat over a much larger area than just one biggie.

Same current must flow through all three seriesed primaries exactly the same as they are in series. (current equal in all parts of a single circuit)..... keeps them matched.

The circuit itself will allow it,backfeed I figure, as the transformers take the 240v forced into the secondaries (240v) ,transform through to the primaries at 60v (20+20+20) This then can be dealt with by the mosfets.

They can pulse them to give us a controlled rectifier and they are connected to the batteries.......

Whether they chose to let you do this  (computer) is another question entirely.... but electrically it is doable  methinks.

Thats as much as i can crystal ball at this distance I'm afraid. I expect boB is the man for the real answers.....


.................oztules

[oztules]

Well, still/ gentle breeze for most of the day so far (2.20pm), but the sun is out, the rain is out, and the wind has come up in fits and starts..... but good testing weather.
The battery bank is mostly full to the brim, so that makes it hard but very useful at the same time.

The sun is out one second, and gone the next... we have variable conditions, and a battery bank that goes overvoltage and back down again, depending on the sun... so we get to test all the boundaries in a short time.

Firstly the   battery was near full at the start.
Turned on three grid tie inverters... total 4.5kw max. With no direct sun (clouded over,) it put out about 900 watts.

When they synced into the inverters output, the current drain from the power inverter went from about 15A battery drain to just positive. Battery voltage rose from 13.7 to 14v per 12 volts (thats how it reports it. For meX4 for 48v)

So now the house was running on pure solar through a solar inverter at 240v50hz. The batteries were gently charging at about 5-10 amps from the solar grid tie output.
...... all was going swimmingly..... it never lasts long does it.....

Now the sun broke through, the battery charger was charging the batteries at over 55amps, and the voltage rose very quickly (as you would expect on full batteries).

At this point I didn't know if I should turn it off or just hang in till something gave out. I figured I could repair it anyway, and we did need to know what will happen at the boundary.

So I waited to see who would give in first..... then click, the charge current dropped negative, but the inverter was still driving the house.... nothing blew up as I had hoped ( but quietly feared). It had been driven over voltage (I suspect) on the output, the inverter dropped out for a fraction of a second, and the grid ties dropped out thinking they had lost grid.... all was well,and the grid ties would return in another few minutes or  so.

I ran out and turned 2 of the grid ties off, now we were back to 1.5kwmax.

The grid tie inverter came back on line, and the current went from discharge to 7-15amps  charge. Our batt voltage stayed around 14v for a bit, the sun went in and were were back to discharge.

So far we have learnt that the inverter will protect itself by an instantaneous shut down continue, and the grid inverters drop out, and the system recovers instantly... so no meltdown in the worst possible conditions.

Next we know it can be quite happy to drift from charge to discharge and all over the place, provided the total load the batteries can put on plus the load is more than the peak output of the solar bank.

We can see that for a bang bang system, we can use the batt voltage for the cut off for grid tie inverters, If it's cloudy and batts are down, voltage will be down, turn on all three. As Batt voltage comes up to 14v... drop off another inverter. If it still climbs, turn off all inverters, and let voltage drop off, and reconnect 1 inverter and test again..... etc etc.


So with a bang bang the system will work, albeit a little lumpy from the ac side of things, but there is no sign of bumps on the ac line.
Grid inverters are soft start so don't stir up trouble with abrupt start........ well there you go.

I used 3 inverters, because they were cheap, but a single one would be fine if it was big, you n=just need to increase the hysteresis to suit the bank size, and chose the grid tie size to suit the load and batteries.... and as much solar as you can garner. Can't have too much.

As I was typing this, the power conditioner gave a single beep. I went outside, to see that the inverter had shaken the grid tie off as it was unable to put the excess power in the bank any more, and so it shut off for a few milliseconds.

The appliances didn't see it, but the power conditioner did. I watched for another 2 minutes, and sure enough the current slowly (1 second or so) went from 10A discharge to 15A charge again... the grid tie had synced in again.

In my case I would only need a single 1.5kw grid tie .. tied to a 8kw array to do what is needed. Come sun up, it would soon be doing it's 1.5kw and stay at 1.5kw all day till night. On bad days, it would still do it.

When the darn new panels turn up we can test the theory fully, but the tests show the possibilities.         

So I would appear I cant kill anything in this setup  regardless of what goes wrong.

This Chinese thing may have too big a idle draw ( poor magnetising characteristic ), but it is tough as nails.

One other thing I found, is that the current control knob for battery charging works when using AUX input power, but is unresponsive to turning when reverse charging. So it looks as though the fets are behaving as expected, and the computer is not controlling the switching  to favour the batteries, but they are just rectifying the excess.

Conclusion. Switching out inverters is easy and as SMA says, is the simple way to do it... and good enough for them is good enough for me....
But to protect the batteries from undue over charging ( which flooded ones love anyway) a dump load (PWM style) will be a nice finish.

Well there it is. It is easy to see how good the SMA system would be if you can control the GTI outputs.... just a walk in the park then.... or slumming it like me.


................oztules
Edit mistakes

[ChrisOlson]

Well, in the old days Xantrex used to tell how their units work right in the Owner's Manual.  So I decided to RTFM.  This is what it says:

The Sine Wave Plus employs a patented inverter design. This design uses a combination of three transformers, each with its own low frequency switches, coupled in series and driven by separate interconnected microcontrollers.  In essence, it is three inverters linked together by their transformers.



By mixing the outputs from the different transformers, a sine wave is produced. This waveform is shown in Figure A-2, “Sine Wave Plus Inverter Output Waveform” on page A–7. Notice the “steps” form a staircase that is shaped like a sine wave. The total harmonic distortion in this sine wave approach is typically 3-5%. The multi-stepped output is formed by modulation of the voltage through mixing of the transformers in a specific order. Anywhere from 34-52 “steps” per AC cycle are present in the waveform. The heavier the load or lower DC input voltage the more steps there are in the waveform.

This type of inverter solves many of the problems associated with high frequency or ferroresonant sine wave inverters. The low frequency method described has exceptional surge ability, high efficiency (typically 90-95%), good voltage and frequency regulation, and low total harmonic distortion.

The inverter runs in two basic formats: as a stand-alone inverter (converting DC to AC), or as a parallel inverter (with its output synchronized to another AC source). In inverter mode, only 60 Hz waveforms are created. As the battery voltage rises, waveforms with progressively fewer steps are generated. More steps are used when battery voltage decreases. Since the battery voltage tends to drop with increased load, the waveform has increased number of steps with heavier AC loads.

The inverter is able to synchronize with other AC sources before connecting it to the AC load. The frequency of the AC source is tracked and the inverter constantly adjusts its frequency to maintain a lock. A normally open contactor is used to parallel the inverter's output and the AC source.

The inverter's power topology is bi-directional. If the waveform created by the inverter has a higher voltage than the paralleled AC source, then power flows from the batteries to the load. If the waveform generated has a lower voltage than the AC source, power flows from the source to the battery.

The various modes of operation use different algorithms for determining the size of the waveform to be created by the inverter. In battery charger mode, for example, waveforms smaller than the AC source are created to cause current to flow into the batteries. This process is fully regulated to provide a three-stage charge cycle. If the level of AC current exceeds the user programmed generator or grid size, and then the inverter will switch to a generator support mode and create waveforms that are larger than the AC source. This causes power to flow from the batteries to the AC loads to prevent overloading of the AC source.

The Sine Wave Plus offers an extremely good efficiency curve. The inverter reaches 95% efficiency at very low AC load levels, which is important because the inverter often spends the majority of the time at the lower power range. The efficiency is maintained over 90% for a wide power range up to the continuous kVA rating. Only when operating at high power levels at or above the continuous kVA rating of the inverter does the efficiency begin to drop off. Since this usually only occurs for short periods of time, the impact may be negligible.
--
Chris

[fabricator]

OZ, by next spring I will hopefully be ready to go big on solar, so hopefully by then you'll have it all doped out.

[Eric L]

But to protect the batteries from undue over charging ( which flooded ones love anyway) a dump load (PWM style) will be a nice finish.

Could it also work to have the grid tie inverters turned on and off by a relay that monitored battery voltage?

To me, it just seems a little scary that you're relying on an overvoltage situation on the inverter's output to trick the grid ties into shutting off. Although it's cool that it's working.

[oztules]

Yes Eric,

Thats exactly how I see it working with dumb GTI's and dumb power inverters.... I like the SMA approach.... it's really versatile smooth and downright cool.

But battery voltage monitoring is the SMA approach for their dumb units too..... and as they seem have cracked this, I figure they have also got very good reason the go that way for the dumb units.


It was DUMB luck that mine worked as it did, and as simply as it did for the testing phase, but I too cringe at the prospect of using the O/V  as a cut off tool permanently.

Some GTI's are remote switched, and this would be first preference, then interrupting  the AC power  the GTI would be next. Using the GTI's mppt to limit the input power to the home inverter from a large array is a very useful current limiter too.


....................oztules

[ChrisOlson]

Well, so the big question is, how can I get it to work with my inverter?  If I could get this to work without the inverter shutting down I got a Jacobs turbine I could use here.  I'd revisit it if somebody told me what the hell I did wrong when I tried it before.

oz - you're the man with answers.  So tell me how I can do this   

My inverter has a grid input (AC1).  If it has power on this it passes thru to the loads and uses excess to charge/maintain batteries.  If grid power is lost it automatically goes to battery backup.

How can I make a standalone grid with my Jake to feed that AC1 input?  There HAS to be a way to do it.  I can't sync the grid tie inverter with the inverter's output and feed it to AC1 because as soon as it "sees" input on AC1 it goes to Grid Support Mode and uses pass-thru.  So now you have a "loop" going where the pass-thru power (really grid tie inverter power) is being used to also feed the AC1 input.  This don't work because it's not stable.

What can I use for a source that will provide a stable 60 Hz 240 split phase "signal" for that inverter to sync with and form a standalone "grid" with it?  If figure if I don't need all the power it produces the turbine has a governor on it and it will just free-wheel up against the governor and no harm done.  It would be just like a standby generator that can generate 10 kW but you only use 2 kW of it.

If the wind stops blowing or the output drops below nominal, the inverter "sees" the "grid" went down and switches to battery backup and off-grid just like we are now.

There HAS to be a way to do this.  I been trying to figure it out for two years.
--
Chris

[hydrosun]

I know that Hugh Piggot has done ac induction hydro into a trace 4024 direct. He may have some ideas on how to do what you want.
Chris

[oztules]

Your inverter does not seem to support grid tie to the output from what I can gleam from you so far.... check with Xantrex maybe it can be programmed for it.


What output does the Jake have....is it a built in grid tie?, or hv ac or......


................oztules