Tell me about ac coupling

[Scruff]

10 year commercial battery? I've seen a few AGMs turn to the wayside after 11 but not consistently. 3 of my 9 died of entropy...kindov half life effect.
I've never seen a flooded die that wasn't killed.

[joestue]

what about trying a thin plate pure lead battery.

lead could be recovered from smelting batteries down yourself. i've done this before but ended up making thermite and burning a hole in my steel container.

other times i was able to get a reaction started but "burn" slowly. i suspect the key might be to just heat the lead oxide waste in a kiln and heat up it up slowly in an inert atmosphere using charcoal to reduce the lead (a solid source rather than using aluminum like i tried) so the reaction can't run away.

i suspect it should be really easy to roll lead sheets yourself, but it may be easier said than done.

[Astro]

First off I have to say, you guys are smart. I can not keep up with the terminology but I am on the same page. Well sort of, I am late to the party and have much building to do before I can get to the experimenting part of it.
But I have already said to my wife, you know our neighbors work different hours, if they each had panels or a mill that provided half a house worth of electricity and since they are not home at the same time, if they shared, they would be set. You guys are obviously calling it ac coupling. I gather you are trying to do it by sending,.... lets call it different signals through the same wire. This seems very doable. I can also see where some people might not like that, so I am just going to not say a whole bunch until I figure out how much to say. All I know is does anybody remember when seti was saying what they could do if they could link up all the little satellite dishes in everyone's yards? Same concept really.
 As for making batteries, I too have been thinking about this. I used to weld poly when I was working on chemical/electrolysis  process of deburring small stainless steel parts. So my thought was to plastic weld up some cells. I have a grasp on the whole thing, but then I got to the part of cost of lead sheets. Since I have a lot of other things to build and acquire before I need batteries, I have only briefly glanced over some numbers. I think if you are going to do this, you had better plan on salvaging your own lead, because I priced sheet lead and very quickly arrived at the conclusion you did, that it is going to be hard to compete with just buying commercial batteries.
I am sure in the quest for AH per dollar a guy could come out ahead, but I think only if you make your own lead sheets. But again I did not dig real deep. Fact I only spent a few minutes checking a couple of thing out.
 I guess it is cool to know that when I get down this project string a little, we are all thinking along the same lines. I look forward to working on these things with you all. I should hopefully be a ways down the project string by spring and I am not doing anything in the winter here anyway (it gets cold). So I won't be working outside putting up anything. I might get some shop testing done before spring, but until I get some serious heat and some insulation blown in above the shop, I won't be out there much either when it is cold.

[clockmanFRA]

Joestue,

I have been practically experimenting with the PLANTE battery components over the past 20 years. A PLANTE battery is all lead plates. I have experimented with lead, zinc, lead, zinc but its a hassle and zinc so easily distorts with any heat.

I have dismantled and had a 'look see' and repaired different standard commercial designs and types.

Most use a lead matrix grid system and imbedded a paste mix of some very unpleasant chemical pastes. The lead matrix is pure lead but only makes up about 1/3 rd of the battery weight, its the lead oxides chemical pastes that make up the rest.

Now i have seen a utube vid, a year back cant find it now,  of a backstreet company in India recycling commercial manufactured lead acid batteries. Crap! man, all very fascinating, but hell life expectancy and breeding a family is not very good in that environment.
They remove the plates and get everything off them back to the lead matrix grid.  The pastes are re-constituted, although this was not fully explained, and in a remade soft putty paste form re applied to the lead matrix grids and allowed to set. Then the battery is re-assembled into new cases.
The other day i received one of these, and was told by a French supplier it was New and high quality etc etc, Just under one sticker/badge, put on by the French retailer, was a stamp mark 'Made In India'.   I tested it with my battery test gear, and it had only 50% of the amperage capacity of a proper new, new commercial battery of the same size and case design. Yep it was one of the India recycled types.

As for just taking a battery and smelting it down. Nope could not recommend that concept.

For a good PLANTE battery i am going to use Uk industrial Lead roof flashing that comes in rolls of various thicknesses and widths. Its not  PURE, PURE lead in the true chemical sense of purity but its acceptable.

I do have smooth rollers to crush down to a specified thickness. Again this thickness depends on hanging lead sheets and seeing them distort and sag over time to optimise for the case length of a completed PLANTE 2 v cell at 50kg, and at the moment best i can achieve is 1.7mm thick, so this area needs a allot of thought.
 Each plate will get a thinnish Polypropylene woven material sock so the plates can be tightly packed into the case so there would be minimal use of Sulphuric acid 15 to 1 mix, and this tight packing would be similar to a true AGM structured cell with the glass matt cell separators. Although not as tight as a commercial AGM as these tend to bulge and crack the cases at the end of there life.
 Then there is the possible need for polypropylene plate comb separators at the plates base and at the top to avoid distortion contact between the main Plus and negative Bus bars.  Most AGM's, the ones i have had apart, cells fail on this as the cell ages it raises 7mm, and there is only a manufactured clearance of 10mm, and that cell is then toast.

Anyways, a fair bit to sort out in many parameters to get to a Do It Yourself PLANTE battery of 50kgs at 2 volts and about 400ah and publish a DIY book.

 Yes I do have chemical fume cabinets and even a chemical room so i can do most stuff on experimentation in a safe manner.

[clockmanFRA]

Ac Coupling is a good thread topic, but we are deviating a bit, so I have started a new thread topic in storage about PLANTE batteries i am developing.

https://www.fieldlines.com/index.php?topic=150445.msg1057077#msg1057077

[Mary B]

what about trying a thin plate pure lead battery.

lead could be recovered from smelting batteries down yourself. i've done this before but ended up making thermite and burning a hole in my steel container.

other times i was able to get a reaction started but "burn" slowly. i suspect the key might be to just heat the lead oxide waste in a kiln and heat up it up slowly in an inert atmosphere using charcoal to reduce the lead (a solid source rather than using aluminum like i tried) so the reaction can't run away.

i suspect it should be really easy to roll lead sheets yourself, but it may be easier said than done.

I posted this elsewhere but will echo it again, NEVER ATTEMPT to smelt down a modern car battery, some of the chemicals released are heavy gasses called arsine and stribine and were used as war gasses to destroy the other sides lungs.

If you want a deeper explanation read here https://castboolits.gunloads.com/showthread.php?138362-Why-Car-Batteries-Are-Dangerous

[joestue]

except that there is no antimony or arsenic in car batteries.

they even tried antimony grid positive plates and calcium grid negative plates in the same battery.. doesn't work. any chemically active antimony plates out on the negative plate and the battery resumes the high parasitic self discharge and high water loss characteristic of lead-antimony grid plates.

[Mary B]

Yes there is, it is alloyed into the lead. The guys over on cast boolits have measured the  lead content using XRF meters...

[Scruff]

There's a lot of Pros to AC coupling, especially when applied to an installation with AC loads.

The most efficient system has no conversion. Hence my little installations on liveaboards are primarily DC Input, DC Output (with AC capability). Ie. DC-Coupled solar to say 24V loads (lights/laptop/audio etc).
If you want a go-anywhere vessel. Do that.

For scalability and integration purposes of most larger installations it's not practical nor beneficial to not use 230V AC Mains so here we accept losses like; I need an extra solar panel to feed the system conversion loss and inverter quiescent...and we have space for this. In fact that space saves more because on a liveaboard where you can only fit two panels efficiency is a must. In a garden it's not such a big deal. Panels are cheaper than efficiency.

This opens us to the potential of acquiring used GTIs with higher output power than equivalently priced DC charge controllers with much larger Voc capabilities for optimising cabling requirements (cost + losses).
These GTIs can be placed anywhere there is mains because they are not required to be local to a battery, so outbuildings can have their voltage droop mitigated by local panels.

To make the ideal system we convert AC -> DC or vice versa the least amount of times. Each time we do we lose power through the heatsinks of the conversion device. Usually 10% but it can be 50% or more depending on the hardware you use to convert this power.

I find it best to use both.

You can make the best anything you want in the world. It's rarely as reliable as 2 good enough ones.

[Scruff]

Another advantage to AC coupling is inverter stacking.

If I DC couple a 60A MPPT to my 24v battery with a 3.5kW inverter downstream I have a 3.5kW power supply.
If I AC couple a 1700W GTI with a 3.5kW battery inverter downstream I now have a peak output of 5.2kW.

[clockmanFRA]

Very nicely said Scruff.

Your paper is very well written, yes a good paper. Might ask you if i can put that in my book please.

But don't forget the main Inverter H Bridge that can backfeed AC turn it to DC and charge/backfeed the batteries. Now that i am getting the hang of the back feeding regime i think i can calm down my direct DC coupled to batteries PV input to only 2.5Kw instead of the present 5Kw. This 5kW DC is on 2 very expensive MPPT controllers and i really only use these as battery charging finishers. Both have there own independent charging regimes and deal with things as required. But are not connected into the main system or themselves. Note, I much prefer the Morningstar Tristar MPPT controller over the Midnight classic 200.

I think for me its the fact that since AC coupling with all my GTI's, the GTI's are the ones feeding my MINI GRID and feeding my power use around all my buildings, so my my main OzInverter just acts as a stable AC voltage, HZ controller and gate keeper.

So once the GTI's input into the MIN GRID my battery use becomes minimum.

The other nice thing about AC coupling to a MINI GRID all you need is the main AC cable installed from building to building, as you would normally with any buildings.   Yea, i normally throw a DC data cable in the cable trench from building to building as well, but over the years as i fine tune my AC COUPLING system, the data cable is no longer used.

I class the OzInverter as a true 6kW output all day if the batteries could cope, however with AC Coupling, i am now back feeding and operating 12Kw of GTI's on the MINI GRID the OzInverter is managing.

Scruff, OzInverter can give 15kW for 20 minutes or so, but i chickened out on any longer as my 1300ah 48v battery bank voltage dropped rather rapidly.

Yes it can do 50kW surges no issue, but 'Oztules' did that on some of his testing procedures, I am very chicken when it comes to that test, so i let 'Oztules' test stand on that one.

Oh yes, the OzInverter is reasonably efficient as it uses only 35w idle power. I do not want to improve this to much as other issues raise there ugly heads regards Choke saturation and inductive feedback problems.  But i do like the possibility of a small EW laminate transformer with huge copper insulated strips wrapped around it to give a non-saturation choke that can be fine tuned.  At present I use and recommend doubled up very large Ferrite cores with a small gap between the core half's, with a minimum of 3 windings of 50mm/2.

With my brute/raw AC Coupled and DC Coupled system i do recommend a minimum of 800ah at 48v battery bank, (possible 600ah) but this depends on the Main Inverter and its backfeeding as the batteries act as a buffer. Also by doing the windings on the toroid to exact voltage readings and number of turns the backfeeding can be limited to stop the backfeeding DC above about 62v.  As most GTI's will knock off at about 254vac. The toroid windings limit us to a Lead Acid 48v battery bank.

 

[noneyabussiness]

just to add a tidbit if not already mentioned,  the beauty of ac coupling is during the day when the sun is out, the ac coupled inverters do most of the heavy lifting with loads.... very efficiently.... the actual inverter doesn't need to work much

[Scruff]

Your paper is very well written, yes a good paper. Might ask you if i can put that in my book please.

You'll have to ask Studer CM. It's their intellectual property. I just happened to agree with it before I read it.

But don't forget the main Inverter H Bridge that can backfeed AC turn it to DC and charge/backfeed the batteries. Now that i am getting the hang of the back feeding regime i think i can calm down my direct DC coupled to batteries PV input to only 2.5Kw instead of the present 5Kw. This 5kW DC is on 2 very expensive MPPT controllers and i really only use these as battery charging finishers. Both have there own independent charging regimes and deal with things as required. But are not connected into the main system or themselves. Note, I much prefer the Morningstar Tristar MPPT controller over the Midnight classic 200.

I only use MS to harndle proper lead charge because no fans (although often HF singy) and specific gravity results. I buy a substantial stock of the MS MPPT orphans that turnip on ebay. This is the thing about buying robust high fidelity equipment. You can rest assured a used one is more often than not right as rain. If you buy at the right price you can sell for the same after you've had your way with it.

If we examine the graph in my previous post we see a system that is extremely well-matched input to load for that particular day at least.

I've added some data points to observe an interesting ratio.



Deploying ball-park-o-meter-scope.

Let's assume Night Load = 2.3 times the Day Load. Because the graph is pretty accommodating about lining up at ~2.3x amplitude if superimposed.
We also observe a ratio of 1:3 AC to DC coupling.
For the following reasons I expect.
AC coupling meets on-demand daytime load profile.
Battery loads are ~2.3 times greater than daytime; to provide this is a DC Coupled Array 3 X size @ 70% efficiency to use at night = 2.1x Equivalent Day Load Supply capability.
For an AC-coupled array to supply the same nighttime power it would need to be 6% larger.

One {extra solar panel / less shadow} per 20 panels.

The benefit of course of DC coupling is the redundancy whereabouts the disappearance of the installation's 230VAC voltage source does not take the battery with it.
For example when the Battery Inverter UnderVoltage Disconnect Trips the GTIs anti-island and hence reduce charge/recuperation opportunity.

What I also draw from the dataset is why are people calling batteries a greener solution for normal grid connected domestic applications? When moving from a 96% efficient system (red) with a lower setup cost and faster return of investment to a 66% efficient system with a battery to pay for too and short-changing the network 30% fossil fuel offset?



The removal of a battery from the system is the best for the network.

A Utility Battery is another matter.

As for people who want to use night rate lecky and day cycle offsetting daytime import?
If night rate = 50% day rate. Night energy = 66% efficient. Battery doubled cost of installation...

So once the GTI's input into the MIN GRID my battery use becomes minimum.

That's what doing it right looks like. I never understand why people focus more on a battery (an inefficiency bucket) than the system input.
(System input = system output) / now  ....that's the proper way.

I class the OzInverter as a true 6kW output all day if the batteries could cope, however with AC Coupling, i am now back feeding and operating 12Kw of GTI's on the MINI GRID the OzInverter is managing.

My powerplant Xtender handles AC coupling fine and DC coupling fine. It's tripped it's own over-voltage threshold by latent backfeeding response in..you know those cold Autumn days when sunbeams are lensing cloud edges like an eyefulla arc weld and the array pummels the system with surges. Hitting two moving targets at instantaneous max output ain't easy. I could increase the Over-Voltage alarm trip. At the moment I've AC coupled one Xtender and DC coupled another.
The LFP is AC Coupled.

It can trip it's own over-voltage with SmartBoost hysteresis too if the threshold is too narrow.
I have yet to see if my 7kVA twinset fair any better managing a system built entirely of other manufacturers' supporting hardware inclusive of 4kW mixed couplings.

My LFP 4kVA is hibernating as of this weekend. The PowerPlant Twinset (one active, one standby) are now handling DC Coupled AC Injection. My SB1700 is on shore power only. I will link it later to backfeed through the Xtender with islanding frequency out of limit trip control for back-up usage.
There seems no point in being concerned about power injection offset & making an even distribution of oversupply between here and April and besides my solar freakin' battery heater refuses to exist on the basis of shenanigans.  There will be no oversupply after the Immersion Diverter for the next 5 months.

The issue I was having is the TS Absorption + Temp compensation voltage was higher than the Inverter Over-Voltage Trip.
The Xtender can be a bit slow to react at back-feeding input surges so bigger batteries are more capable of absorbing these power pulses.

Yes it can do 50kW surges no issue, but 'Oztules' did that on some of his testing procedures, I am very chicken when it comes to that test, so i let 'Oztules' test stand on that one.

50kW..Woowee!!  It'd take me a coupla minutes to find a suitable test load.



It's not a speed limit...it's a challenge.

Oh yes, the OzInverter is reasonably efficient as it uses only 35w idle power.

Yurp. My jaw did hit the floor the first time you said that.

That's quite impressive for a lump that size.
I'd say that puts you in the top 3.

[Scruff]

I think AC coupling charge control/back-feed limitation methods is a bigger topic than AC coupling.

Certainly backfeed limitation is very interesting because if you can use a GTI on the output of any bog-standard battery inverter and regulate it by never backfeeding this would be an extremely cost effective large MPPT solution. That could drive isolated battery chargers as opportunity loads.

The problem with AC diverters whether they admit it or not is that they have a minimum spill to grid threshold. They always bias some backfeed and their reliance on CTs to regulate diversion is very inaccurate coupled with RF signalling...it's too sketchy to couple to a non-hybrid inverter. However....

I could use my powerplant inverter as current source for off grid power with AC coupled solar backfeeding to another inverter that is the voltage source. The current between them is sent to the AC diversion load - the requisite backfeed bleed escaping the diverter can be used by a load on the voltage source upstream of the CT. This load could be triggered by the GTI side of it's own anti-islanding relay (solar only secret circuit)

This allows for a lottov headroom, even moreso if the voltage source is a hybrid with a partially depleted battery.
I reckon it's overengineered.

You could simply spec a DC diversion load that's ample to handle a GTI if the battery inverter is hybrid and the DC diversion load was as useful.

Concept conceived to address how to retain grid-tied system functionality while on backup.

[clockmanFRA]

Thanks Scruff for doing a reply to your STUDER doc.

After reading your reply and amendments to the Studer doc, what you are saying makes sense in a real practical manner.

I think at the moment 'empirical evidence' on real site applications where you do not throw buckets of money at it, is very limited. So reading your comments about smaller scale AC Coupling projects is very enlightening.   PS, No offence for your small system intended.


"I think AC coupling charge control/back-feed limitation methods is a bigger topic than AC coupling."

Oh yes, totally agree.  I have tried to simplify it with my 4 main methods for our practical use, but to be honest each machine and each system would be very different, and i do not think the 'big boys' manufacturers can grapple with the complex issues involved, so hence there package installs.

As i said earlier, i never realised i could use 'Voltage sag' between installations to my benefit.
 
 All GTI's manufactured across the World are preprogramed with differing countries settings to combine with that country's utilities grid. So I am surprised NO ONE YET has come forward with a little box of simple tricks that is an attachment on the GTI to control it, without opening the GTI box or adding a data feed, when using AC Coupling on a MINI GRID system.
Yes its a big call, But hay!, I live in hope!.

Regards battery size...... Is for me one of the main reasons for going the AC Coupling route in the first place, as i said earlier for our 48v 1300ah battery bank having 12Kw AC coupled and backfeeding we have no issues with the OzInverter as the gate keeper. I have reduced this down to 800ah on tests still okay but at below 600ah the batteries are becoming sensitive and the GTI's are 'knocking of' very frequently. So there is a correlation between AC Coupling input Kw and actual battery size.  But i feel sure that 'SCRUFF' has a better handle on that.

PS. PLANTE battery is in case design stage, 2v 550ah, Now PVC case, but for loading and 50kg weight testing and stress etc.

UMM not many one of PVC fabricators around, plenty in China i see, so no choice but to manufacture fabricate the case myself, to achieve a usable and dismantlable cell design, ie take the top of.
 Machined and manufactured with Polypropylene but not PVC, but can use some very good PVC solvent adhesives with primers and drill tap A4 stainless steel 4mm bolts and plastic welding on the seams.
Even have a good reasonably priced plastic ridged sheet stockist here in France, so 10mm thick PVC grey sheets will be ordered.

Best get on, as servicing No 1 wind turbine at the moment, repainting the blades now.   

[Mary B]

Your paper is very well written, yes a good paper. Might ask you if i can put that in my book please.

You'll have to ask Studer CM. It's their intellectual property. I just happened to agree with it before I read it.

But don't forget the main Inverter H Bridge that can backfeed AC turn it to DC and charge/backfeed the batteries. Now that i am getting the hang of the back feeding regime i think i can calm down my direct DC coupled to batteries PV input to only 2.5Kw instead of the present 5Kw. This 5kW DC is on 2 very expensive MPPT controllers and i really only use these as battery charging finishers. Both have there own independent charging regimes and deal with things as required. But are not connected into the main system or themselves. Note, I much prefer the Morningstar Tristar MPPT controller over the Midnight classic 200.

I only use MS to harndle proper lead charge because no fans (although often HF singy) and specific gravity results. I buy a substantial stock of the MS MPPT orphans that turnip on ebay. This is the thing about buying robust high fidelity equipment. You can rest assured a used one is more often than not right as rain. If you buy at the right price you can sell for the same after you've had your way with it.

If we examine the graph in my previous post we see a system that is extremely well-matched input to load for that particular day at least.

I've added some data points to observe an interesting ratio.



Deploying ball-park-o-meter-scope.

Let's assume Night Load = 2.3 times the Day Load. Because the graph is pretty accommodating about lining up at ~2.3x amplitude if superimposed.
We also observe a ratio of 1:3 AC to DC coupling.
For the following reasons I expect.
AC coupling meets on-demand daytime load profile.
Battery loads are ~2.3 times greater than daytime; to provide this is a DC Coupled Array 3 X size @ 70% efficiency to use at night = 2.1x Equivalent Day Load Supply capability.
For an AC-coupled array to supply the same nighttime power it would need to be 6% larger.

One {extra solar panel / less shadow} per 20 panels.

The benefit of course of DC coupling is the redundancy whereabouts the disappearance of the installation's 230VAC voltage source does not take the battery with it.
For example when the Battery Inverter UnderVoltage Disconnect Trips the GTIs anti-island and hence reduce charge/recuperation opportunity.

What I also draw from the dataset is why are people calling batteries a greener solution for normal grid connected domestic applications? When moving from a 96% efficient system (red) with a lower setup cost and faster return of investment to a 66% efficient system with a battery to pay for too and short-changing the network 30% fossil fuel offset?



The removal of a battery from the system is the best for the network.

A Utility Battery is another matter.

As for people who want to use night rate lecky and day cycle offsetting daytime import?
If night rate = 50% day rate. Night energy = 66% efficient. Battery doubled cost of installation...

So once the GTI's input into the MIN GRID my battery use becomes minimum.

That's what doing it right looks like. I never understand why people focus more on a battery (an inefficiency bucket) than the system input.
(System input = system output) / now  ....that's the proper way.

I class the OzInverter as a true 6kW output all day if the batteries could cope, however with AC Coupling, i am now back feeding and operating 12Kw of GTI's on the MINI GRID the OzInverter is managing.

My powerplant Xtender handles AC coupling fine and DC coupling fine. It's tripped it's own over-voltage threshold by latent backfeeding response in..you know those cold Autumn days when sunbeams are lensing cloud edges like an eyefulla arc weld and the array pummels the system with surges. Hitting two moving targets at instantaneous max output ain't easy. I could increase the Over-Voltage alarm trip. At the moment I've AC coupled one Xtender and DC coupled another.
The LFP is AC Coupled.

It can trip it's own over-voltage with SmartBoost hysteresis too if the threshold is too narrow.
I have yet to see if my 7kVA twinset fair any better managing a system built entirely of other manufacturers' supporting hardware inclusive of 4kW mixed couplings.

My LFP 4kVA is hibernating as of this weekend. The PowerPlant Twinset (one active, one standby) are now handling DC Coupled AC Injection. My SB1700 is on shore power only. I will link it later to backfeed through the Xtender with islanding frequency out of limit trip control for back-up usage.
There seems no point in being concerned about power injection offset & making an even distribution of oversupply between here and April and besides my solar freakin' battery heater refuses to exist on the basis of shenanigans.  There will be no oversupply after the Immersion Diverter for the next 5 months.

The issue I was having is the TS Absorption + Temp compensation voltage was higher than the Inverter Over-Voltage Trip.
The Xtender can be a bit slow to react at back-feeding input surges so bigger batteries are more capable of absorbing these power pulses.

Yes it can do 50kW surges no issue, but 'Oztules' did that on some of his testing procedures, I am very chicken when it comes to that test, so i let 'Oztules' test stand on that one.

50kW..Woowee!!  It'd take me a coupla minutes to find a suitable test load.



It's not a speed limit...it's a challenge.

Oh yes, the OzInverter is reasonably efficient as it uses only 35w idle power.

Yurp. My jaw did hit the floor the first time you said that.

That's quite impressive for a lump that size.
I'd say that puts you in the top 3.

50mph road? Challenge accepted! LOL I used to drive a dragster at 160mph, and a stock car on dirt... plus many a time dodging cops as a teen by going down field tracks at a high rate of speed LOL I like driving fast!

[Scruff]

  PS, No offence for your small system intended.

Challenge accepted! LOL

[Bruce S]

50kW..Woowee!!  It'd take me a coupla minutes to find a suitable test load.



It's not a speed limit...it's a challenge.


50mph road? Challenge accepted! LOL I used to drive a dragster at 160mph, and a stock car on dirt... plus many a time dodging cops as a teen by going down field tracks at a high rate of speed LOL I like driving fast!

Mary B;
SOoooo I'm not the only one who read that, that way  . As a former ridge runner, Pro-stock and a once in a lifetime chance to drive on the Indy track.

That Pic looked like (except for the speed sign) something leading up to the 1-lane bridge back home

Bruce S