Sunny Island 5048U charging strategy

palausystem · April 2014

I'm ignorant when it comes to how other controllers work. I do know that the SI is not the first choice for off-grid use. But I'm trying live with what I got.

Right now I have Boost set at 2.5v for 2.5 hrs. The SI goes from float into Boost when SOC drops to 70% or when 30% of nominal capacity has been used (added over time). So for us that means 2-3 times a week. I then try to do a 5 or 6 hr absorb once/week to get it into the high nineties or 100. Boost often brings it into the low nineties. So I'm trying to make sure I get a Full charge once/week.
One irritating thing about the SI is that it will not do an automatic Full charge like it's supposed to. If you have Full set for 5 hrs at 2.5v and every 7 days, it will look back over the week and say, "oh, you already met that criteria with your Boost charges so I'll just skip it this time and reset the counter. Wouldn't want to overcharge!" According to SMA that's just the way it is.

So if I want to mimic a "regular" off-grid controller, (absorb daily?) I suppose I could reduce the Boost parameters to something really minimal, and then set Full (with adjusted settings) to happen every day, thus bypassing Boost altogether. Then do a once/week 100% charge like I'm doing now (using the manual EQ with adjusted settings).

As far as automatic operation goes, how do other controllers handle things? My goal is to have as little interaction as possible while still getting a decent life-span from the battery.

Speaking of lifespan, our Surrette S-530's are 6 years old and I'm trying to get through next winter with them. For the first 3 or 4 years of their life they spent most of the time in the low to mid 80's SOC with rarely any kind of full charge. Our installer told us the SI would baby our batteries for long life. When I discovered the Full charge he had set for every two weeks was not happening, I began to learn more. Eventually with the help of this forum I began to play with the settings. Now that we are in the market for new batteries, I'd better have a clear idea of how to handle the SI.
Thanks...

David and Laura · April 2014

Re: Sunny Island 5048U charging strategy

We also have a SI-based system, and have similar charging problems preventing our battery SOC from going much above 90%. SMA support keeps advising us just to "give it time", so we've been waiting until we have a long string of sunny days before trying to adjust our settings.

BB. · April 2014

Re: Sunny Island 5048U charging strategy

The thought is, usually, once the battery is >90% state of charge at least every few days--You are doing fine. Going to 100% every day is hard on the battery and inefficient (much of the energy goes to gassing the battery and to heat).

How is your water usage... It seems if you need to water the cells about every 2 months--You are doing fine. If you need to water every month or more often, probably over charging. If you need to water every 6 months or less, probably under charging (although, there are some batteries that are designed to use less water like the newer Trojan RE series).

-Bill

Cariboocoot · April 2014

Re: Sunny Island 5048U charging strategy

SMA's Sunny Island is designed for backing up a GT system. It's not really suitable for straight off-grid use where charging is supplied by AC coupling from GTI's. Don't expect it to work ideally in such an application. Another case of the wrong equipment for the job.

There's a work-around: get a charge controller and enough panel to do the finish charging. You won't need full current capacity for this; just observe the current rate when the SI stops charging. Of course if you've no more room (or money) for additional panels this won't work.

Cariboocoot · April 2014

Re: Sunny Island 5048U charging strategy

BB. wrote: »

The thought is, usually, once the battery is >90% state of charge at least every few days--You are doing fine. Going to 100% every day is hard on the battery and inefficient (much of the energy goes to gassing the battery and to heat).

No, it isn't. This is a myth perpetrated by people with incorrectly designed systems. The majority of off-gridders can and do manage daily 100% charging and have battery banks that last for many, many years.

Sometime "100%" charging is confused with "130%" charging; incorrect parameters that have too high a Voltage or too long an Absorb time or automatic EQ when none is needed. That is hard on batteries.

zoneblue · May 2014

Re: Sunny Island 5048U charging strategy

Funny you should mention this about Sunny island. I just happened to be looking at stephenvs data feed here:

https://xively.com/feeds/75889/

Click SOC, then set the range to a month. The chart suggets that that system hasnt hit 100% all month, yet if you study the solar watt curves its clear that charge is terminating most days. I dont know whats going on there, maybe something to do with SMAs interesting "only absorb if DOD below X%" algorithym.

Stephen knows SMA gear, so maybe he can help you guys.

zoneblue · May 2014

Re: Sunny Island 5048U charging strategy

Cariboocoot wrote: »

No, it isn't. This is a myth perpetrated by people with incorrectly designed systems. The majority of off-gridders can and do manage daily 100% charging and have battery banks that last for many, many years.

Sometime "100%" charging is confused with "130%" charging; incorrect parameters that have too high a Voltage or too long an Absorb time or automatic EQ when none is needed. That is hard on batteries.

And if your daily cycle is shallow, full daily absorbs will then create grid erosion, as ive heard you explain it?

Cariboocoot · May 2014

Re: Sunny Island 5048U charging strategy

zoneblue wrote: »

And if your daily cycle is shallow, full daily absorbs will then create grid erosion, as ive heard you explain it?

Yes; a balanced system operates in a "sweet spot" where the discharge is not so shallow as to be a waste of battery capacity and not so deep as to significantly shorten life. Remember that the numbers deal with "average use" so there is bound to be day-to-day variations for both discharge and charge. The trick is to make sure those variations aren't outside the practical range.

stephendv · May 2014

Re: Sunny Island 5048U charging strategy

Cariboocoot wrote: »

SMA's Sunny Island is designed for backing up a GT system. It's not really suitable for straight off-grid use where charging is supplied by AC coupling from GTI's. Don't expect it to work ideally in such an application. Another case of the wrong equipment for the job.

This is just plain wrong. The Sunny Island was designed right from the start as a 100% off-grid inverter. You CANNOT connect it to the grid in any way in Germany where it was designed. If you want grid backup, then they have a different product for that "Sunny Backup".

The SI's charging strategy is more sophisticated than other charge controllers. So it's a bit like asking how to make your Porsche 911 go slower, because all the other cars are slower than yours

Palausystem, the surrettes seem to need higher and longer absorbs than other batts, so you can try setting the boost charge to a longer value. 2.5V is high enough for boost, IMO but perhaps some of the surette owners will chime in. If you set the Full time to a longer value, won't you force it to do a Full charge as it'll be unlikely to have satisfied the parameters during boost?

stephendv · May 2014

Re: Sunny Island 5048U charging strategy

zoneblue wrote: »

Click SOC, then set the range to a month. The chart suggets that that system hasnt hit 100% all month, yet if you study the solar watt curves its clear that charge is terminating most days. I dont know whats going on there, maybe something to do with SMAs interesting "only absorb if DOD below X%" algorithym.

I'm cheating because I'm not using the SI as a charger, instead using a custom raspberry pi + midnite classic. Chose this route because it was cheaper and more efficient than SMA's AC coupling solution. I don't even know what the SI's Full charge is set to, it may be too long a value so it never sees the classic charge the batteries that much and therefore never reaches 100% SoC. Suppose I should change it

The SoC reading on the SI is only ever about +-10% accurate so I take the readings with a pinch of salt. Just completed a short EQ today, batteries are at 1.29 on the hydrometer, yet the SI says 95% SoC.

Cariboocoot · May 2014

Re: Sunny Island 5048U charging strategy

stephendv wrote: »

This is just plain wrong. The Sunny Island was designed right from the start as a 100% off-grid inverter. You CANNOT connect it to the grid in any way in Germany where it was designed. If you want grid backup, then they have a different product for that "Sunny Backup".

The SI's charging strategy is more sophisticated than other charge controllers. So it's a bit like asking how to make your Porsche 911 go slower, because all the other cars are slower than yours

Palausystem, the surrettes seem to need higher and longer absorbs than other batts, so you can try setting the boost charge to a longer value. 2.5V is high enough for boost, IMO but perhaps some of the surette owners will chime in. If you set the Full time to a longer value, won't you force it to do a Full charge as it'll be unlikely to have satisfied the parameters during boost?

No you are just plain wrong.
The SI was design to provide back-up power in an outage and work perfectly with the Sunny Boy GTI's, which it does. I never said it was meant to be connected to the grid because it's not a grid-tie inverter. It's charging system does not "work better". It works differently and that is not the same thing.

As an off-grid inverter it's less-than-adequate, and the proof of that is in this thread. I believe I described this system as "crackers" originally and was lambasted for it.

Well let's see: it's too expensive, inefficient, and it just plain doesn't do the job. So that's good, is it?

At least it's not one of those lousy Xantrex/Schneider/Conext XW things. The SMA equipment actually works without needing upgrades or having to ignore numerous flaws.

Too many people fall into the prejudice that because they have a certain brand that is the one and only and absolute perfect unit. I fix systems and see them all and know better than to defend any one brand as being the ultimate. Outback's inverters are significantly better than other OG units but they have drawbacks as well, some of which are really annoying.

stephendv · May 2014

Re: Sunny Island 5048U charging strategy

Cariboocoot wrote: »

No you are just plain wrong.
The SI was design to provide back-up power in an outage and work perfectly with the Sunny Boy GTI's, which it does.

It was designed as an off-grid inverter. It may be popular as a backup power inverter in the US, but it's not legal to use it that way in Germany. There are some good reference sites where it's a 100% off-grid inverter, e.g.: 110kW system here: http://www.sma.de/en/products/references/eigg-island.html
and 140kW: http://www.sma.de/en/products/references/atoll-reao.html

I don't blindly trumpet SMA kit. As I just said above, I think their AC coupling idea is unnecessarily expensive and inefficient for small-medium sized off grid homes like mine. Which is why I went the DC route with a midnite. Their smaller inverters are also too expensive compared to the likes of outback and Victron, which why I don't recommend SMA for < 5kW systems.

But that doesn't mean that if you already have an SI + SB AC coupled system that it won't work. It will and does.
Since its boost, full and eq parameters are completely configurable, I think you're confusing a non-optimal configuration with a fundamental flaw in the equipment.

Cariboocoot · May 2014

Re: Sunny Island 5048U charging strategy

As an off-grid inverter the Sunny Island's charger makes no sense. As a GT back-up it is ideal.

Maybe those German engineers aren't as smart as we've been led to believe.

stephendv · May 2014

Re: Sunny Island 5048U charging strategy

Cariboocoot wrote: »

As an off-grid inverter the Sunny Island's charger makes no sense. As a GT back-up it is ideal.

Maybe those German engineers aren't as smart as we've been led to believe.

Or maybe there's more than 1 way to charge a battery. The main difference with SMA's method is that they want to address overcharging by not doing an absorb at all if the batteries are sufficiently charged. Some newer charge controllers are starting to include this function too, like the Studer Vario series: http://www.studer-inno.com/?cat=mppt_solar_charge_controller which lets you skip absorb if the batts are above a certain voltage. So the idea is not unique to SMA.

Midnite and outback achieve a similar goal with their end amps settings, although they go to absorb every day even if they can end it quickly. Morningstar's MPPT has a simple fixed absorb time, so has no way to deal with this.

Cariboocoot · May 2014

Re: Sunny Island 5048U charging strategy

It would be better if you could skip Absorb if the batteries were above a certain SOC. That sounds like a challenge for MidNite! They may be able to program that in to work in conjunction with the WhizBand Jr.

It also would be better if the current during Bulk could be ramped up slowly, mimicking panel charging. Outback's ramps up (takes on loads first, then increase charge for whatever portion of input power is leftover) but does so fairly quickly. This would be a more complex process.

One of the faults of all built-in inverter-chargers is the lack of Bulk clocking. This is probably why SMA put the "no Absorb" function in, as the others simply assume you want 2 hours of Absorb no matter what (as opposed to a good solar charger which limits Absorb time to Bulk time, max Absorb, and/or End Amps). If the built-in chargers operated more like the good controllers it would be a huge improvement.

But they don't because that costs money. In the off-grid world it would be considered a largely unnecessary function because the charger is usually only used on non-sunny days and powered from a gen making maximum current from the start more efficient in terms of gen use. If you're there and watching you can manually shut off charging before Absorb goes full time and/or if current drops low (from lack of need). If you're not there and watching you are at the mercy of the programming for your particular unit. You could tell your charger to have zero Absorb, but then if you have a stretch of bad weather and need it you'll need to tell it otherwise.

I have not done an Outback Flexnet DC install, but I suspect it may offer these sort of advantages if you can find the right settings to tweak.

stephendv · May 2014

Re: Sunny Island 5048U charging strategy

Cariboocoot wrote: »

One of the faults of all built-in inverter-chargers is the lack of Bulk clocking. This is probably why SMA put the "no Absorb" function in, as the others simply assume you want 2 hours of Absorb no matter what (as opposed to a good solar charger which limits Absorb time to Bulk time, max Absorb, and/or End Amps). If the built-in chargers operated more like the good controllers it would be a huge improvement.

The SI _does_ offer a programmable timed absorb, in fact 2 independent settings ("Boost" and "Full") where you can set both the absorb time and the voltage. It's just that the Boost won't run everyday if the SoC is high enough. And the "Full" runs every X days, also configurable.

boB · May 2014

Re: Sunny Island 5048U charging strategy

Cariboocoot wrote: »

It would be better if you could skip Absorb if the batteries were above a certain SOC. That sounds like a challenge for MidNite! They may be able to program that in to work in conjunction with the WhizBand Jr.

This is definitely one feature that will be added as soon as we get a chance.

Re-Bulk based on SOC has been the intention for years but not until the WB Jr.
was available which it now is so that can happen.

All sorts of possibilities are possible knowing the battery current.

Not sure what you mean by ramped up bulk charging current ?
Don't know what gain that gives you. Maybe you could expand on that a bit ?

Cariboocoot · May 2014

Re: Sunny Island 5048U charging strategy

boB wrote: »

Not sure what you mean by ramped up bulk charging current ?
Don't know what gain that gives you. Maybe you could expand on that a bit ?

It's a phenomenon many of us off-grid have noticed but haven't been able to document explicitly: batteries seem to charge "better" from solar than from AC powered chargers. Apparently it's a function of not having full current available first thing in the morning (as opposed to the AC chargers starting out with 100% power capacity). It seems the charging is better if it spends time at a lower current to begin with and increases gradually to full power.

The chargers in the Outback due this in seconds, which does not duplicate the effect seen with the solar that occurs over hours. Not an efficient use of generator power, but there may be some balance available there. I'd envision a charging profile where the current percentage goes up as the Voltage climbs, but never exceeding the peak limit.

If only we could accurately define just what it is that makes it "better". SG comes up the same either way, but the slow charge seems to last longer (which doesn't actually make sense). Perhaps it is better at desulphating the plates.

boB · May 2014

Re: Sunny Island 5048U charging strategy

OK, I see what you mean. Well, maybe just turning down the current limit on the inverter/charger would get it closer to how it charges as the morning goes towards noon on a PV system ?

Cariboocoot · May 2014

Re: Sunny Island 5048U charging strategy

boB wrote: »

OK, I see what you mean. Well, maybe just turning down the current limit on the inverter/charger would get it closer to how it charges as the morning goes towards noon on a PV system ?

Yes, but then you'd have to turn it up gradually as charging progresses; you wouldn't want to charge at 5% all day when you could charge at 10% and get it done in a few hours.

Just something to think about for MidNite's future inverter-charger with the separate power and control modules that allow input of 12 to 48 Volts via buck-converter and interchangeable, stackable 500, 1000, and 2000 Watt power increments.

Well we can dream, can't we?

palausystem · May 2014

Re: Sunny Island 5048U charging strategy

In response to Stephendv's question about Full charge duaration vs Boost duration...
I could try raising the Full absorb time to see if it will actually kick in. If I do raise it to say 8 hrs and it kicks in, I will have to terminate it manually because 8 is too much. And so the automatic factor is moot. Really what I think would work on average is a 2.5 hr Boost which would bring it into the high 80's to low 90's every few days, then a 5 or 6 hr Full charge once a week to get it to the high 90's/100. Then I could feel free to ignore things for a few weeks at a time during Spring and Summer. But as I said, there seems to be no way to get this to happen automatically with the SI.
I'll keep playing around with it though.
Everyone here seems to be in love with the midnight classic.

stephendv · May 2014

Re: Sunny Island 5048U charging strategy

palausystem wrote: »

I could try raising the Full absorb time to see if it will actually kick in.

Or try raising the Full absorb voltage instead, that way Boost will never satisfy the Full criteria.

palausystem wrote: »

Everyone here seems to be in love with the midnight classic.

I think at a stage midnite were considering modifying the classic so that it would follow the SI's charging program.

Cariboocoot · May 2014

Re: Sunny Island 5048U charging strategy

Everyone loves the Classic because it incorporates so many features we've been longing for. The WhizBang Jr is just the latest and greatest addition to a unit that already has "Waste Not" to make better use of otherwise unrealized solar power and "Follow Me" for accurately coupling multiple Classics as well as "HyperVoc" protection for cold climates.

Very good design. The best so far.

palausystem · May 2014

Re: Sunny Island 5048U charging strategy

Yes, I think I will try that. Just looking at the thread talking about how the Surrettes like the higher voltages anyway.

palausystem · May 2014

Re: Sunny Island 5048U charging strategy

Actually, I forgot - I tried that awhile ago. I set Boost at 2.45v for 2 hours and Full at 2.5 for 5 hours and it still wouldn't kick in.
So now I'm going to try Boost at at 2.5 and Full at 2.58 and see what happens.

stephendv · May 2014

Re: Sunny Island 5048U charging strategy

palausystem wrote: »

Actually, I forgot - I tried that awhile ago. I set Boost at 2.45v for 2 hours and Full at 2.5 for 5 hours and it still wouldn't kick in.
So now I'm going to try Boost at at 2.5 and Full at 2.58 and see what happens.

Hmmm. I really expected that to work. Have you tried setting Full to every day so that it's faster to test theories about how to get it to kick in?

midijeep · August 2017

With the SMA Sunny Islands you need to change the Float voltage to a voltage that keeps the SOC where you want it e.g. 95% , 100% etc... I have been battling this for a while and finally won the battle. You will use more water but if 100% SOC is what you are shooting for then change the Float voltage. With my 48 voltage battery bank (38 cells) I need to keep my Float voltage at 1.55 to keep the SOC at 100%, which draws around 200 watts or 4 amps.

Sunny Island 5048U charging strategy

Comments

Categories