Helping another off-gridder

Surfpath

Ok, so helping a buddy who thinks his AGM batteries are on their way out after 4 years. He has been relying upon his installer for help to date. Professes to not be a solar expert. I'm going to take a look at his system myself next week. This is what he told me about his system so far.


"10 Evergreen ES - 180 panels

1 Sunny Boy 1700 Inverter
1 Sunny Island 3324 Inverter / Charger
6 Lifeline AGM GPL-4DL Deep Cell Batteries with a capacity of 14KWH"

I know that each of his Lifelines are rated at 210 amp hrs (& 12V).

His area gets lots of sunshine (6 good hours) for 9 months a year with less in the rainy season (2-6hrs depending on the day). His panels are all on the western side of his hip roof (South is optimal). He'll invariably have some shading issues in the morning. I believe he is underpanelled, but by how much? Also, I think those Sunny Islands are hyper VOC chargers. So I predict that he'll have all his panels in one string.

Any thoughts/tips regarding his specs? What should I be thinking about?
Cheers, SP

Cariboocoot

Re: Helping another off-gridder

So there are six 210 Amp hour 12 Volt batteries? That would be 630 Amp hours on a 24 Volt system.
Normal array for that would be near 2kW. He's got ten 180 Watt panels so it's pretty close.

The Sunny Island is an off-grid inverter-charger, not really a charge controller. It sounds like it is AC coupled to the Sunny Boy 1700.

All panels pointing West isn't great. He could be seriously low on production that way.

I'd check the battery bank configuration first; make sure it is wired optimally. Then check each battery string in turn; charge it, let it sit for at least 3 hours (overnight is better), see how it holds Voltage. Look for differences in Voltage between batteries. Problem with AGM: can't stick a hydrometer in. So you have to go by Voltage and Voltage retention.

I'm not sure if the SB has any readout on it for Watt hour production, but look at that if it does. And then of course look at Watt hour consumption. You know what happens when there's a deficit there!

He might want to think about augmenting the solar with some more panel and an actual charge controller to pick up power early in the day. Even if its just enough to offset his use before the main array starts producing it will extend the life of the batteries.

westbranch

Re: Helping another off-gridder

Another thing to check is that all the connections from the PVs to the CC and then to the battery are all tight, a loose one can cause a lot of trouble, even minutely loose. Copper has the ability to 'flow' after first tightening leaving a seemingly looking good connection loose.

Do you have a clamp-type ammeter?

stephendv

Re: Helping another off-gridder

Surfpath wrote: »

I believe he is underpanelled, but by how much?

That's impossible to tell without knowing his loads and where he is.

First things to check would be the charging settings recommended by the battery manufacturer, and then check that the Boost, Full and EQ settings on the sunny island match those.

Surfpath

Re: Helping another off-gridder

stephendv wrote: »

First things to check would be the charging settings recommended by the battery manufacturer, and then check that the Boost, Full and EQ settings on the sunny island match those.

Ok, thanks. Here are the GPL-4DL Charging Voltages per the Concord Lifeline site:
Bulk Charge Voltages: 14.20 - 14.60 -
Absorbtion Charge Voltages: 14.20 - 14.60volts
Float Charge Voltages: 13.10 - 13.40

Checking the voltages/cables and inquiring about his loads sound fine.

Coot, I think I get what you mean by: "The Sunny Island is an off-grid inverter-charger, not really a charge controller". I found a copy of the Sunny Island manual [URL="http://"]here[/URL]. I now understand it's an AC-coupled system:

"AC coupling joins the various power sources on the AC side. With a PV system, this technique requires using more than one inverter since DC sources of power are still present and each power source must be converted to AC before they are joined.

The ... PV array feeds through an SMA America Sunny Boy string inverter. The output of the Sunny Boy inverter is then routed to the AC load center—just as in a batteryless, grid-tied system. This system also uses two SMA Sunny Island inverters to convert 48 VDC from batteries into 120/240 VAC power that also feeds the AC load center.

With this AC coupling setup, the Sunny Island inverters provide AC voltage to the AC load center, effectively “tricking” the Sunny Boy inverter into thinking grid power is available and enabling it to send out power from the array. This power can then be used for powering AC loads that are running and/or be pulled back through the Sunny Island inverters to charge the batteries." - Justine Sanchez from Homepower Mag.

So what's the maximum amount of array wattage allowed for this system (or should I be thinking in amps - I see : "A maximum of 56 A should flow through the Sunny Island's AC input"). It's so different than my Outback.

BTW, my buddy is currently completely off grid (however, sounds like his SMA gear would allow for future GT, great).

stephendv

Re: Helping another off-gridder

Surfpath wrote: »

Bulk Charge Voltages: 14.20 - 14.60 -
Absorbtion Charge Voltages: 14.20 - 14.60volts
Float Charge Voltages: 13.10 - 13.40

In the Battery section of the menu, there will be the 3 charge settings for Boost, Full and EQ. They should be within those ranges, would also be good to know the time setting for each as well as the interval for Full charging, I think it's Days Between Full, or something like that. The default setting for Boost charging for flooded lead acid batteries is quite high and if this is the same for AGM then that could be quite a bad thing (tm).

Surfpath wrote: »

So what's the maximum amount of array wattage allowed for this system (or should I be thinking in amps - I see : "A maximum of 56 A should flow through the Sunny Island's AC input"). It's so different than my Outback.

It will be limited by 2 things: the size of the Sunny Boy (1700W) and the maximum charging current of the Sunny Island, if that's 56A x 24V is roughly 1300W. So assuming losses, array size seems fine (relative to the SMA kit). Whether that's enough array for their loads is still unknown. There's also no harm in installing a larger array to produce more during cloudy weather, just that its output will be limited by the Sunny Boy and charge going to the batteries will be limited by the Sunny Island.

Surfpath

Re: Helping another off-gridder

stephendv wrote: »

There's also no harm in installing a larger array to produce more during cloudy weather, just that its output will be limited by the Sunny Boy and charge going to the batteries will be limited by the Sunny Island.

Hi Stephendv thanks for the advice.

I was also hoping for a shortcut to set a minimum voltage cut-point on my buddy's Sunny Island. Ie. a point where the inverter would shutdown when voltage got too low, something like 24.20v (50% SOC).

On my outback's mate it's easy to set this in the ADV settings. Is there a similar shortcut for the Sunny Island?

inetdog

Re: Helping another off-gridder

One warning to you Surfpath:


Although both the Sunny Boy and the Sunny Island take their input from solar panels, you CANNOT connect both the Sunny Island and the Sunny Boy in parallel to the same set of panels. You will have to make a guess, I mean calculation, as to how to split the power between them and divide your array into two sub-arrays.
The rule is that unless they communicate and coordinate with each other (and even then not always) you cannot connect two MPPT devices in parallel to the same set of panels. They will fight for the lion's share of the panel output and will end up with a less than ideal panel output and maybe unstable division of power between the two devices.

stephendv

Re: Helping another off-gridder

inetdog wrote: »

Although both the Sunny Boy and the Sunny Island take their input from solar panels, you CANNOT connect both the Sunny Island and the Sunny Boy in parallel to the same set of panels. You will have to make a guess, I mean calculation, as to how to split the power between them and divide your array into two sub-arrays.

That's not how the Sunny Island and sunnyboy system works, the SI is just a plain old battery based inverter/charger, it has no input for solar. All the panels are connected to the sunny boy, and the sunny boy AC out and the sunny island's AC out are then connected together.

stephendv

Re: Helping another off-gridder

Surfpath wrote: »

I was also hoping for a shortcut to set a minimum voltage cut-point on my buddy's Sunny Island. Ie. a point where the inverter would shutdown when voltage got too low, something like 24.20v (50% SOC).

Yep, it's there in the Battery settings, under Battery Protection I think. The 3324 is a bit dated, so not sure if the setting is activated based on SoC (using its internal SoC meter), or voltage. Newer models only use soc.
Best to check the manual for that model on how it's implemented, on the 5048 model there are 3 different settings: standby at night, off but comes on every 2 hours during the day and finally completely off.

One of the problems with AC coupling is that the sunny island is the only way to charge the batteries, so if it shuts down completely because of low SoC, and then the next day there is full sun, the batteries won't charge unless someone physically goes and resets it. So be very careful about what you set that final SoC limit to. This is why they've implemented the 3 stage SoC shutdown, where the middle setting is: shutdown, but turn yourself on every 2 hours to see whether there's solar power available.

There are also other ways to do this using the internal relays for load shedding. E.g. connect non-essential loads to relay 1 and then program the load shedding to kick in at 60% SoC. And of course auto-generator start with relay 2 to avoid getting the soc down that low.

As I said, things have changed between models so best to check the manual: http://www.rpc.com.au/pdf/SMA_SI3324_4248_user_manual.pdf

inetdog

Re: Helping another off-gridder

stephendv wrote: »

That's not how the Sunny Island and sunnyboy system works, the SI is just a plain old battery based inverter/charger, it has no input for solar. All the panels are connected to the sunny boy, and the sunny boy AC out and the sunny island's AC out are then connected together.

Thanks for the correction.

What the does mean is that either the Sunny Island batteries are charged only from an AC-input charger or from a separate CC. If there is a separate CC to keep the battery bank charged, it still must not be connected to the same panel array as the Sunny Boy.
If the only charging source for the Sunny Boy's battery bank is AC from the grid, then there is no potential conflict on the DC side, But I think that you may have the possibility of a tail chasing situation, under high load, of having net power taken from the joint SI/SB output to be used for the SI to charge its own batteries.

stephendv

Re: Helping another off-gridder

inetdog wrote: »

What the does mean is that either the Sunny Island batteries are charged only from an AC-input charger or from a separate CC.

Nah, that's where the AC coupling comes in. The Sunny Island is bi-directional, i.e. when the sunny boy + PV produce more power than there are loads on the AC side then the SI absorbs that power and uses it to charge the batteries. In a 100% AC coupled system the SI is the single point of control and of charging/discharging the battery.

Many other inverter/chargers behave the same way when a GTI is attached to their output, the Outback's and Xantrex XW will both charge a battery if a GTI is injecting power on the AC output side. A problem with this is that when back-feeding the batteries this way the inverter/charger can't regulate the power, it just charges the battery with whatever the GTI is pushing in. The SI + SB system solves this problem by changing the output frequency slightly, and having the Sunny Boys change their power output based on the frequency.

In the SMA AC coupling view, the AC bus is the central part of the system, not the batteries, with the SI being the single point of control which sets the output AC frequency. You can then attach sunny boys with PV or windy boys with wind turbines all to this AC mini-grid. It's a nice idea for large distributed systems, but a bit inefficient and expensive for small home based systems.

Surfpath

Re: Helping another off-gridder

stephendv wrote: »

Yep, it's there in the Battery settings, under Battery Protection I think. The 3324 is a bit dated, so not sure if the setting is activated based on SoC (using its internal SoC meter), or voltage. Newer models only use soc.
Best to check the manual for that model on how it's implemented, on the 5048 model there are 3 different settings: standby at night, off but comes on every 2 hours during the day and finally completely off.

One of the problems with AC coupling is that the sunny island is the only way to charge the batteries, so if it shuts down completely because of low SoC, and then the next day there is full sun, the batteries won't charge unless someone physically goes and resets it. So be very careful about what you set that final SoC limit to. This is why they've implemented the 3 stage SoC shutdown, where the middle setting is: shutdown, but turn yourself on every 2 hours to see whether there's solar power available.

There are also other ways to do this using the internal relays for load shedding. E.g. connect non-essential loads to relay 1 and then program the load shedding to kick in at 60% SoC. And of course auto-generator start with relay 2 to avoid getting the soc down that low.

As I said, things have changed between models so best to check the manual: http://www.rpc.com.au/pdf/SMA_SI3324_4248_user_manual.pdf

Stephendv,
That's awesome - exactly the advice I needed. Thanks.

I've had to delay traveling to the site for another week, but I will now go there a little more prepared. Mostly I have been asked to confirm that his batteries are indeed on their way out, and to look at ways to improve his PV harvest (reorient or increase panels, perhaps).
-SP

Surfpath

Re: Helping another off-gridder

I visited my Buddy's site last night
In addition to the Sunny Island 3024, he's got a Sunny Boy SB1700. I tried to scroll through the LCD menu on each unit. Overall, pretty difficult to understand what I was looking at (admittedly I'm an Outback user). So I did not do much.

His batteries are t.o.a.s.t. I saw 24.6v at 8pm. Apparently the system now shuts down each day in the early morning b/c of low voltage. The Batts have lasted 4-4.5 years, so not too bad for semi-beginning users.

His loads are about 3.3-4.0 kwh/day. As noted before, his panels face West (and he has a semi obstructing hill to the East) so his production is best from 11am to 6pm. Warm temps year round.

I quickly googled the SB1700 and see:

"Recc maximum PV power 2050 Wp" - - "Max AC power 1700 W"

Does this mean that, regardless of the voltage nameplate on his bank (ie 24v), he is limited to a max of 2050watts of panels (he currently has 1800w)?
Perhaps if he adds a smaller array (3-4 hundred watts?) on the Eastern side, he will stay within specs throughout the day (but increase his chances for an earlier 'bulk'). Is adding a few panels an easy thing using the Sunny System?

Also, I confirmed his bank is comprised of 3 strings of 210amp hr 12v lifelines = 660 amp hrs. What's a better config for his new bank?

My gut feeling is two parallel strings, each string has four 6v batteries (like these L16RE-A's at 325 amp hrs = 650 AmpHr 24v bank).

Cariboocoot

Re: Helping another off-gridder

Surfpath wrote: »

I quickly googled the SB1700 and see:

"Recc maximum PV power 2050 Wp" - - "Max AC power 1700 W"

Does this mean that, regardless of the voltage nameplate on his bank (ie 24v), he is limited to a max of 2050watts of panels (he currently has 1800w)?

Yes, because it is AC coupled. The maximum output of the Sunny Boy is 1700 Watts, period. It does not know or care about the size of the battery bank on the Sunny Island. This would be the same if it was a simple grid-tie system with no SI involved.

Perhaps if he adds a smaller array (3-4 hundred watts?) on the Eastern side, he will stay within specs throughout the day (but increase his chances for an earlier 'bulk'). Is adding a few panels an easy thing using the Sunny System?

Array input in two different orientations, so long as the Vmp matches, should not cause difficulty and should extend the usable hours of output. If there is "too much" panel (within reason) on it the extra power won't be realized.

Extra panel may not be necessary, however, if the SB1700 can reach it's maximum power: divide 4kW hours by 1.7 and you get 2.4 hours equivalent good sun. You say there's 6+ hours of sun available, but perhaps not ideal angle. It should be able to produce the 4kW hours in that time span though.

Also, I confirmed his bank is comprised of 3 strings of 210amp hr 12v lifelines = 660 amp hrs. What's a better config for his new bank?

Fewer strings of larger capacity. The SI in this case is 3kW so 630 Amp hours is pretty good for it. Possibly two strings of 320 Amp hour L16's instead? Only needing 4kW hours per day is good: on a 24 Volt system that's 167 Amp hours consumed - less than 200 even with conversion & losses. It would dial in around 400 Amp hours minimum @ 24 Volts.

My gut feeling is two parallel strings, each string has four 6v batteries (like these L16RE-A's at 325 amp hrs = 650 AmpHr 24v bank).

I agree that would be suitable.

stephendv

Re: Helping another off-gridder

Surfpath wrote: »

In addition to the Sunny Island 3024, he's got a Sunny Boy SB1700. I tried to scroll through the LCD menu on each unit. Overall, pretty difficult to understand what I was looking at (admittedly I'm an Outback user). So I did not do much.

Yeah, you'd need the manual to make sense of the codes There wouldn't be anything to check on the sunny boy, the important battery charging settings are on the sunny island in the Battery menu. Boost, Full and EQ voltage settings, time periods and intervals.

Surfpath wrote: »

Perhaps if he adds a smaller array (3-4 hundred watts?) on the Eastern side, he will stay within specs throughout the day (but increase his chances for an earlier 'bulk'). Is adding a few panels an easy thing using the Sunny System?

The SB is a normal grid tie inverter, so if you add panels to the other side they'd need to be at the same voltage as the existing array. The SB's voltage range goes up to 320V, so if the current array is at 240V for example, then you'd have to buy enough panels to match that. E.g. with 30V grid tie panels in the 240W range you'd need 8 of them, i.e. 1920W. Or remove panels from the existing array to bring the voltage down, but be sure that it stays within the SB's input range.

Alternatively, install the extra 400W of panels using a regular charge controller on the battery. You'd also have to wire up a cheap shunt on the battery negative so that the SI's knows about the extra charging source.

EDIT: You could even just use a cheap PWM charge controller and grid tie panels to keep costs down.

Surfpath

Re: Helping another off-gridder

stephendv wrote: »

The SB is a normal grid tie inverter, so if you add panels to the other side they'd need to be at the same voltage as the existing array. The SB's voltage range goes up to 320V, so if the current array is at 240V for example, then you'd have to buy enough panels to match that. E.g. with 30V grid tie panels in the 240W range you'd need 8 of them, i.e. 1920W. Or remove panels from the existing array to bring the voltage down, but be sure that it stays within the SB's input range.

Hmm, can he just add a few more panels to the same string? And locate these panels on the Eastern side of his roof?

So he currently has 10 Evergreen ES 180's: (ISC 11.84, IMP 10.53, VOC 21.3, VMP 17.1). This means he currently has a total of 213 VOC on his single string of panels (Caveat: I presume it's one string because with Sunny you can do that - why do a different way).

Coot said "Array input in two different orientations, so long as the Vmp matches, should not cause difficulty and should extend the usable hours of output".

So could he add 2 additional panels with a similar VMP on the same string? If he were to source the same type of panels as the Evergreens that would give him a total of 255.6 VOC (below the 320 max). His total array size would grow to 2,160 watts (ie. more than the 1850watts input rating that the SB 1700 has. But not all those panels will be "on" at the same time.

There is a nice discussion related to adding panels to a SB1700 here on the Whirlpool site. I am still trying to absorb it.

Cariboocoot

Re: Helping another off-gridder

You don't want panels on the same string in two different aspects: the current will be uneven and the bypass diodes will work overtime.

You can parallel a string of equal Voltage rating and have it face in a different direction.

If the existing array is not at its maximum I'd try increasing that first, providing there's room.

And as Stephen said a different array on its own controller going directly to the batteries is also an option. How much PV could be fitted to the East/South? Even 30 Amps coming in earlier could be a big help.

Surfpath

Re: Helping another off-gridder

Cariboocoot wrote: »

You don't want panels on the same string in two different aspects: the current will be uneven and the bypass diodes will work overtime.

You can parallel a string of equal Voltage rating and have it face in a different direction.

If the existing array is not at its maximum I'd try increasing that first, providing there's room. (Coot: I think he's at the SB1700 limit for PV in with 1800 watts already there)

And as Stephen said a different array on its own controller going directly to the batteries is also an option. How much PV could be fitted to the East/South? Even 30 Amps coming in earlier could be a big help.

OK, got it.

The East has a lot of room (with 3000watts+ potential). There is actually a tiny south facing section (2 panels can fit there, maybe 3). But then again it would be difficult to create a string of equal voltage rating there. So, a string of equal voltage on the East side is probably best.

Since the incoming max for the SB1700 seems to be 1850watts (right?), and he already has 1800 watts (10 panels) facing West, perhaps the solution is to move 5 of those panels to the East side on a second string. Also since my buddy would like to increase his array (I think it's b/c he has struggled at times with his present set up and he wants to ensure success), perhaps he can add a panel to each string? He would then have 1080 watts on each side.

(pause)
Hold on. When reading the Whirlpool thread (here it is again) I am still a little confused. It says on page 2 for the SB1700 you need "a Voc of at least 180V to get your inverter started." This is also reflected in the SB1700 manual which says "Voltage range (extended operating range) 180 V ... 265 V."

So, given his panel specs (ie. 21.3 VOC per panel), he would actually need 9 panels on each string? Six of his panels would only add up to 127voc. That sounds a little crazy, but perhaps its because his highish wattage panels come with low VOC (they are essentially 12v panels??). Panels like mine (220w with ~35 VOC) would have been better.

Still a little unsure.....

ps. also, from the SB1700 manual...it says "max DC input current....12.6A." Would two strings exceed this if his current panel Isc is 11.84? I believe with strings you have to add amps together (while the system voltage remains the same as for each string). An example are my 9 panels which are in 3 strings resulting in an array voltage max of around 105volts (ie 35voc x 3), but about 28-33 amps (7.81 ISC per panel).

Cariboocoot

Re: Helping another off-gridder

Okay it sounds like the SB is at it's limit already. If the string on it has an Isc of 11 and the I max is 12 you clearly can't double it up. I would expect the input max on a 1700 Watt GTI to be about 2200 Watts.

That makes the choice of array & charge controller the one to go with.

But the existing array may not be right. The Vmp (not Voc) of the array has to be in the operating range of the GTI. If this is 180 Volts for start-up and the panels are 'standard' 12 Volt units with a Vmp around 17-18 the string must be ten panels minimum. It may well be that his array is operating in the bottom of the V in range, which would reduce over-all output. If possible, I'd add two more identical panels to the string. This will not alter the current but will increase the Vmp.

These panels sound like Evergreens of old: 200 Watts (based on 1800 Watts from nine panels) with a Vmp around 17 (based on Voc of 21.3 and W/Imp) and Imp around 11.8. That will be difficult to match, and it is essential the Imp is very close; a panel with lower current rating will "choke off" the rest of them reducing the array's power. Unfortunately very few panels have that high a Imp these days.

You are correct that these panels were a poor choice for this application. Not much can be done about that now.

Surfpath

Re: Helping another off-gridder

Cariboocoot wrote: »

..But the existing array may not be right. The Vmp (not Voc) of the array has to be in the operating range of the GTI. If this is 180 Volts for start-up and the panels are 'standard' 12 Volt units with a Vmp around 17-18 the string must be ten panels minimum. It may well be that his array is operating in the bottom of the V in range, which would reduce over-all output. If possible, I'd add two more identical panels to the string. This will not alter the current but will increase the Vmp.

These panels sound like Evergreens of old: 200 Watts (based on 1800 Watts from nine panels) with a Vmp around 17 (based on Voc of 21.3 and W/Imp) and Imp around 11.8. That will be difficult to match, and it is essential the Imp is very close; a panel with lower current rating will "choke off" the rest of them reducing the array's power. Unfortunately very few panels have that high a Imp these days.

You are correct that these panels were a poor choice for this application. Not much can be done about that now.

So, given the Isc limit of the SB1700, two strings are not feasible & the focus should be on how to expand his current string.

This is where I don't understand the SB1700 Specs in the manual, because on p. 37 it lists a "Max OCV of 400v" and then it states a "Max. input power Pdc of 1850 W."

If two additional panels are added to his existing string of 10 panels his Vmp (17.1 * 12) will rise to 205v, but his panel wattage will go from 1800 to 2160w. In your mind even if this increases the Pdc (?) is this still OK?

Aside 1: With Isc and Vmp being so variable in panels, how can a wattage limit be set?, Surely it's a current limit? (ok light is going on) Maybe it's a "string" wattage max?

Aside 2: As you point out, finding a similar sized high wattage 12v panel will be challenging. These are 6 year old Evergreens.

Cariboocoot

Re: Helping another off-gridder

The Maximum Open Circuit Voltage of 400 is not a worry. That's one good thing.
How they are determining the maximum input power is another issue. Is it an absolute or an average? You'll have to ask SMA that. Typically panels put out about 80% of their rating (panel heats up as sun angle improves - rarely is rated Wattage achieved). So 2200 Watts (eleven 200 Watt panels) would output 1760 to the inverter, but is that within the SB's tolerance level? Again a question for SMA.

I thought there were nine 200 Watt panels? No? It's ten 180 Watt panels?

The Wattage limit is based on IN/OUT: the maximum output of the inverter is fixed so that places a limit on the input. As with MPPT charge controllers there's no sense over-paneling to the extent where most of the time the 'extra' panels do nothing. But there are also input limits for both Voltage and current as you have seen, and all must be taken into account. Exactly what the operational difference will be is a matter for the particular inverter. 12 Amps (max current) @ 180 Volts (min Voltages) is 2160 Watts, for example. 12 Amps @ 260 Volts is 3120 Watts but the inverter will only output 1700 even so.

The panels are already at the current limit, but the Voltage is low for being in 'ideal' MPPT range for the inverter. It's one string, so there's no way to re-arrange the existing PV for more optimum performance. The panels aren't available anymore, and finding similar ones may be impossible. I see three options here, neither of which are very good:

1). Parallel a couple of new panels on one 'end' of the existing string to allow the current pass-through while increasing the Voltage (possible problems including exceeding Wattage maximum, choking of current if one of these panels fails, overloading of bypass diodes if shading occurs).

2). Change out all panels for new ones that are better suited. NO ONE wants that expense. Even selling the old panels won't recoup much.

3). Leave as-is, suffering the loss and improve the system function otherwise. The new array to the East & MPPT controller should take care of this, but will not actually increase the efficiency of the SB1700.

If SMA says it would be okay to add panels to the string to increase the Voltage at existing current level then that is possible, but the shortage of similar panels still makes it an unlikely prospect.

I think the problem here is that the array Vmp is falling below the minimum when the sun angle is less than perfect, so even with long hours of daylight the GTI doesn't actually produce during much of that time. Is it possible to observe its output over the course of a day and see when/how much it is actually producing?

Oh I suppose we could add options 4 and 5: replace the SB1700 with micro inverters matched to pairs of panels or replace it with a MPPT charge controller direct to the batteries since that's about all it's doing anyway. A charge controller's minimum Voltage would be >battery level, which means 35 Volts not 180. Even a Classic 200 wouldn't take the Voc of the string as-is so it would have to be split and paralleled. And then you'd probably still want to add panels facing the other way to start charging earlier.

Cariboocoot

Re: Helping another off-gridder

Addendum to changing the SB1700 to a charge controller:

1800 Watts on an MPPT controller feeding a 24 Volt system should provide about 58 Amps peak current, enough for 580 Amp hours. This is near the existing battery bank size - just a little under.
At 25% DOD such a battery bank would provide just about 3kW hours AC per day.

This option may not be as silly as it first appears. Especially if a similar Vmp array can be added facing the other way (connected to the same charge controller) to add charging capacity and allow for a slightly larger battery bank. Roughly $600 (plus re-wiring costs) for a Classic vs. what to try and salvage a badly designed AC coupled system?

stephendv

Re: Helping another off-gridder

Since the existing panels are off-grid panels, you could group them in strings of 2 and connect all of them to a cheap PWM charge controller, then add some new panels to make a new array for the SB1700 on the other side of the roof. Common grid tie panels have a Vmp of 30V so you'd only need 6 new panels to meet the voltage requirements of the SB.

Surfpath

Re: Helping another off-gridder

stephendv wrote: »

Since the existing panels are off-grid panels, you could group them in strings of 2 and connect all of them to a cheap PWM charge controller, then add some new panels to make a new array for the SB1700 on the other side of the roof. Common grid tie panels have a Vmp of 30V so you'd only need 6 new panels to meet the voltage requirements of the SB.

Hi Coot, good info. Yup, he has ten 180 watt Evergreens currently. I can now see that a new array to the east with its own controller is probably the best solution.

I got a new bit of info. The owner said that shortly after the system was first installed there was a period of very cloudy weather, and his system was struggling to keep up. The installer supplemented the house array by putting up a temporary trailered array (my guess, eastern facing), that had it's own 'readout', but that it was "plugged into the house AC" (?not sure how this last part worked - maybe it's a Sunny thing).

So the concept of adding additional panels through another array with a separate controller (controller or inverter?) is not a new one.

Stephendv I see what you say about using the old panels to make 5 strings of 2 on an additional PWM controller. That way the new Grid Tie panels would meet the Sunny Boy's 180V minimum, and his overall system would be more efficient. Though, he certainly could create this new array with a set of GT panels (32-35 Volt range) on a PWM controller and leave the old string the way it is.

Either way adding a new PWM array sounds like the way forward. I guess the main thing is that charge points for the PWM controller would have to be the same as what is programmed into his Sunny Island.

I think I finally understand the Sunny Boy/Island system: Using PV power the Boy creates an AC 'Grid.' The Island charges the batteries with this 'grid' power as well as 1) inverts stored battery power for house AC needs and/or 2) uses the 'grid' power to directly supply house AC needs, thereby avoiding the Dc battery loop.

Am I on the right track?

This leads me to the next question how much power can his system (I guess the SI 3024) make. With the Sunny system I am not sure because of this AC coupled flexibility.

Cariboocoot

Re: Helping another off-gridder

Is there enough space for adding a whole new array to power the SB1700?

I can't find any matches to the apparent specs on those Evergreens. Do you have the exact numbers for them? 180 Watts with an Imp >11 means the Vmp will be around 17 at best, so ten in series will not even make the SB's 180 Volt minimum most of the time. In essence much of the array's power generating potential is wasted because the GTI won't even turn on if the input Voltage is too low.

Given the array size, it would be cheaper and more efficient to divide it in two and put it on an MPPT charge controller. Two rows of five in series with a Vmp around (5 * 17) 85 - not ideal for a 24 Volt system but viable and should be fairly easy to achieve.

If you make five parallel strings of two in series on a PWM type controller the Voltage would be fine but you will need a combiner box with breakers/fuses to handle five strings. That is an extra expense. The current output would be slightly less too.

Either of these choices would better utilize the existing PV with no worries about trying to find a match. The SB1700 would then either be re-arrayed or removed and sold. Six GT style panels (up to 300 Watts) would be minimal: eight 230 Watt units would probably be better.

I don't doubt that they could roll a trailer full of panel & another GTI in to supplement during cloudy conditions. That should have been a red flag that the system design wasn't right.

stephendv

Re: Helping another off-gridder

Surfpath wrote: »

Stephendv I see what you say about using the old panels to make 5 strings of 2 on an additional PWM controller. That way the new Grid Tie panels would meet the Sunny Boy's 180V minimum, and his overall system would be more efficient. Though, he certainly could create this new array with a set of GT panels (32-35 Volt range) on a PWM controller and leave the old string the way it is.

Either way adding a new PWM array sounds like the way forward. I guess the main thing is that charge points for the PWM controller would have to be the same as what is programmed into his Sunny Island.

Yes, both charging sources the SI and the PWM controller should have the same charge settings. Grid tie panels are usually cheaper than off-grid panels, which is why I suggested using the existing panels for the PWM off-grid controller, then buying new and cheaper grid tie panels for use with the SB. But if there's no price difference then you might as well leave the existing array as is, and add new off-grid panels to the PWM.

Surfpath wrote: »

I think I finally understand the Sunny Boy/Island system: Using PV power the Boy creates an AC 'Grid.' The Island charges the batteries with this 'grid' power as well as 1) inverts stored battery power for house AC needs and/or 2) uses the 'grid' power to directly supply house AC needs, thereby avoiding the Dc battery loop.

Am I on the right track?

Close All grid tie inverters MUST see an existing grid before they will start producing power. So it's actually the Sunny Island that creates the grid, the SB then sees this grid and decides to inject power into it. The SI then either charges batteries, or supplies loads from the batteries if there's a shortfall. Since the SI's output and the SB output and the loads are all connected together, all loads that run during sunshine hours will run from power directly produced by the SB. Which is slightly more efficient than the typical DC arrangement of Charge controller + battery inverter.

BUT battery charging is more inefficient, because power from the panels is converted to AC in the SB (6% loss), then converted back to DC again when it goes through the SI's battery charger. And then converted back to AC again when it comes back out of the SI to supply the loads.

So the efficiency of directly running loads when the sun's out is great for applications like offices, industry or factories where the loads match sunlight better. But in off-grid homes the "worst" times are winter nights and cloudy winter days. During those marginal times you really want your batteries to be as efficient as possible, and that's where the AC coupling idea doesn't work so well.

Surfpath wrote: »

This leads me to the next question how much power can his system (I guess the SI 3024) make. With the Sunny system I am not sure because of this AC coupled flexibility.

Since the SI and the SB's output are connected together, in theory it can make 3kW (SI) + 1.7kW (SB) = 4.7kW in ideal theoretical conditions. But only 3kW is stable and doesn't depend on the sun shining.

Surfpath

Re: Helping another off-gridder

stephendv wrote: »

So the efficiency of directly running loads when the sun's out is great for applications like offices, industry or factories where the loads match sunlight better. But in off-grid homes the "worst" times are winter nights and cloudy winter days. During those marginal times you really want your batteries to be as efficient as possible, and that's where the AC coupling idea doesn't work so well.


Since the SI and the SB's output are connected together, in theory it can make 3kW (SI) + 1.7kW (SB) = 4.7kW in ideal theoretical conditions. But only 3kW is stable and doesn't depend on the sun shining.

So AC coupled systems suffer a little on those cloudy, rainy days when you beg for every MW of PV. No doubt, coupled with a Western facing array, his batteries had a bit of a fight during those times.

I now see why they call them "Sunny" systems.

Thanks to you and Coot for helping me understand this quite different system. The owner is now getting a quote for a new bank and a new Eastern array.
Cheers,
SP

Surfpath

Re: Helping another off-gridder

Update: It's taking his installer a little while to give him a quote. So I was asked to put something together.

With the assistance of this great forum, I suggest that he add a second, smaller, array of grid-tie type panels to the Eastern side of his roof, connected to a new, smaller, MPPT charge controller which is then directly connected to a new FLA battery bank.

His new panels will provide good bulk charging in the mornings, leaving his ‘old’ West Array to still provide afternoon absorb/float charging. Aside: For this latter array he could also look for one more 12v type panel to better fit the Sunny Vmp minimum (180v)

Otherwise his new total array will be 1800 + 750 = 2550 watts.

This should keep his new 740 amp/hr 24v battery bank happy in his local, relatively sunny conditions: Here is the formula to calculate the 10% charge rate:
24 volts * 740 AH * 1/0.77 panel+controller derate * 0.10 rate of charge = 2306 watts array for a 10% rate of charge

Controller Purchase
Midnite Kid 30A MPPT

Panel Purchase
Three “grid tied” panels like these Sonali 250’s, (60 cell panels) total of 750 watts (Vmp 31.32, ISC: 8.45, IMP: 7.98 ). Given that his record low is 20 degrees c, should be fine for the Kid to handle.

Wire Size: Given an ISC of 8.45 and a run of about 50 feet, 10 awg should be fine (15amp max per the AWG table). No fusing necessary, right?

Battery Bank Purchase
I recommend 8 Trojan L16RE-B Batteries (370 amp hr) = 740 amp hr bank, In 2 strings of 4 batteries each (24volts, 740 amp hrs).

Grounding the Panels: Midnite SPD 115.
Sound about right?:cool:

stephendv

Re: Helping another off-gridder

I think an MPPT controller is overkill, because he can just use a simple and cheap PWM with those grid tie panels all in parallel. Although since they'll be parallel with a PWM he'd need some fusing

BB.

Re: Helping another off-gridder

I would stick with the MPPT type controller. Vmp 30 volt panels are just too low of voltage for charging a 24 volt battery bank.

No fusing for the PV Wiring and you will have about a 1 volt drop--Just fine for a 90 volt Vmp array.

You might want to put a switch/circuit breaker on the PV wiring if the owner ever wants to turn off the array to service/reboot the MPPT charge controller. One of those "nice to have options".

-Bill