My Experience Networking 2 Charge Controllers and Sizing Battery Bank

[Canine]

I had purchased one Blue Sky Energy 3024i thinking it was enough for the panels on the roof; two 285 watt, 60 cell panels were connected in parallel with a total of 570 watts. When I looked at the chart for the 3024, this CC will take 540 watts of 36 cell panels on a 12 volt battery bank, or 400 watts of 60/72 cell panels on a 12 volt battery bank, or 800 watts of 60/72 cell panels on a 24 volt battery bank. I had looked at the 24 volt battery bank column then purchased only one 3024i charge controller when I should have purchased two CCs. Well, the power output was less than I expected. Then as I was reading the manual, I noticed my mistake. It was chopping off 170 watts (maybe more?) and allowing a max of 400 watts worth of power through. Bummer! I was also worried about taxing the CC to the point of burning it up, so I bought another 3024i. These are made to be networked together to act as one charge controller. Furthermore, if one panel gets shaded or a large bird poops on it, the second stays at full power.

I hooked it up and wow! Lots of power. It's no where near double the power, but it feels like it. That extra 170 watts made a huge difference. Where I am parked now, late in the day one panel gets shaded. When I had just one CC, it made a noticeable drop in power. Now it drops only in half. In the winter getting as much solar as possible is so important. Running fewer amps by running 10 gauge wires from each individual solar panel likely reduced voltage drop as well, which increase power to the batteries.

I almost put this paragraph in its own thread, but it's too interconnected with the panels to do that. Figuring out how many battery amp hours I had kind of tricked me, but I have 4 UPS (unlimited power supply) AGM batteries at 140 a/h a piece for a total of 560 a/h. I figured this was a good balance to the panel wattage, and it is during the summer, but not in the winter. Definitely not with the 400 watts I had been getting from the panels. Regarding the battery bank, I need only 400 a/h to comfortably do all my electrical stuff, so I took one battery out (now 420 a/h) to see if charging a smaller battery bank was faster. Turns out it is faster. 50 a/h into a 420 a/h bank is faster than 50 a/h into a 560 a/h bank. Then when the second CC came in, the charge times are soooo much quicker. Three hours of charging to get into float mode is now down to maybe one hour of charging to get into float mode using the same electrical loads on similar days. I can heat (not boil) water for my oatmeal and heat water for the shower then by the time I leave, it is usually in float mode! Even on partly cloudy days! Using the microwave and heating water using a hot pot with 420 a/h is no problem. I was thinking about putting the other battery back in, but this set up is working so well I can't bring myself to change it. I like bringing my batteries to a full charge as often as possible and this set up allows me to do that the most often.

With the original set up the batteries didn't always get fully charged every day. Now there hasn't been one day the batteries haven't been fully charged. On days when I expect snow to cover the panels, I use propane to heat my water, so I do make adjustments. Cloudy days give me much less grief now. I wish I had been more detailed and written down specific info, but know that the difference is considerable.

P.S. Customer service is literally the best I've ever experienced and I've had some great service in the past. Once you are part of the Blue Sky family, you are treated like family. Don't believe me? Try it. I wish I had more questions so I could visit them more. lol

 

[Weight]

I question that smaller bank charges faster than larger bank, with the same use from each. Use 400 ah, replace 400 ah. Larger banks will not drain to as low a percentage and should be easier to recharge than a bank drained nearer to 50%. Your dual panel with individual controllers is a great way to use that much wattage. Thanks for your post.

 

[frater secessus]

Sounds like your system is definitely happier.  

I'm going to dig down a bit for the possible benefit of onlookers and newcomers.  

It was chopping off 170 watts (maybe more?) and allowing a max of 400 watts worth of power through. Bummer! I was also worried about taxing the CC to the point of burning it up, so I bought another 3024i.

400w isn't far off from what one is likely to get with 570w of panel:  570w x .75 derating = 430w.

Running more panel than the controller is rated for is called overpaneling.  It has benefits and drawbacks but Blue Sky says:

Current limit should be thought of as transient PV over current protection and not a normal mode of operation. If

the [controller] enters [the controller's] current limit on a regular basis it is possible that excessive PV power is connected. Confirm
that the PV modules do not exceed ratings shown in the Selecting PV Modules section.

so you are wise to have made a change.  Morningstar, on the other hand, says overpaneling is fine and even gives guidance on how to do it.

Overpaneling can cause excess heat, particularly unloaded in late Absorption when an mppt controller moves the panels toward Voc in order to reduce current.  My understanding is greater difference between Vpanel and Vbatt causes more heat when bucking down.  This increase in heat can be reduced by running one's big loads in late Absorption and Float when the system has excess power.

I have 4 UPS (unlimited power supply) AGM batteries

Unlimited power batteries would be awesome;  unfortunately the U stands for uninterruptible. 

Three hours of charging to get into float mode is now down to maybe one hour of charging to get into float mode using the same electrical loads on similar days.

There is a saying that it takes one woman 9 months to make a baby, but two women cannot make a baby in 4 1/2 months. 

I mention this because it is impossible to speed up Absorption;   the battery bank wants what it wants and that's usually at least 2 hours to accept the last bits of current .  As sternwake once said, you could hook up a nuclear power plant to the batteries and Absorption wouldn't happen any faster.

So I wonder if the controller is entering Float prematurely.  Or maybe it looks at historical data (like Victron?) and is making the right call.  If Float commences at 1hr, one could watch the current going into the battery at 45mins and see if current is nearing C/200 or whatever the battery manufacturer advises.

On days when I expect snow to cover the panels, I use propane to heat my water, so I do make adjustments. Cloudy days give me much less grief now. I wish I had been more detailed and written down specific info, but know that the difference is considerable.

Very cool!   It's a great feeling to tweak a setup and get a nice return.  

Once you are part of the Blue Sky family, you are treated like family. Don't believe me? Try it. I wish I had more questions so I could visit them more. lol

I don't have the kind of money it takes to get adopted into that family. 

 

[frater secessus]

Use 400 ah, replace 400 ah.
...
Larger banks will not drain to as low a percentage and should be easier to recharge than a bank drained nearer to 50%.

I agree with the 400Ah part.  [We might point out for newbies it actually takes a bit more than 400Ah to replace what was used, but that's not the point Weight is making]

And I agree that larger banks could get through Bulk faster.   But I think the time advantage would be due to greater throughput rather than the actual depth of discharge. 

Example:

• we need to replace 100Ah
• manufacturer specifies, say, a C/5 charging rate
• C/5 charging of 100Ah bank = 20A.  5hrs for Bulk.
• C/5 charging of 200Ah bank = 40A.  2.5hrs for Bulk, assuming we can produce that 40A.
• Absorption duration the same

Let me know if I am missing something obvious.  I am getting a late start on my first cup of coffee.

 

[Canine]

Weight, you are right if the charging source can increase the amount of amps it is delivering. If the charging amps stay the same, charging a smaller bank is faster. Since I can now put out more amps with two CCs, charging the larger bank should be relatively fast like you said.

frater secessus, You are right. I did make a mistake when stating my changes in performance. I had my original settings higher than they should have been. The change was: Absorption originally at 14.8 lowered to 14.4 and Float 13.9 to 13.5. I specifically remember the absorption, but can't quite remember what the float was, but it was higher. Now that my settings are where they should be, I am getting into float mode sooner because the distance to the finish line is shorter. I am embarrassed that I overlooked that- thank you for the correction. It's important to have accurate information; misleading anyone on here doesn't help anyone.

As an addition the panels were running about 20 amps through the wires in parallel which made about 3% voltage drop. Now that each panel has its own wire, the voltage drop is only about 1.5%. This translates into about an 8 watt power increase. Not much, but I'll take it.

So far the batteries that came from the UPS system have been performing very well. We'll see how long they last. Next time I do plan on getting new ones that are specifically designed for my application. I picked this current set up for $600 when they were 3 years old, so if they last about 4 years in my possesion, then I will have broken even. 2 1/2 years to go. They may not last that long, but we'll see. I'm not going to lose my shorts on them at least.

Having separate, networked controllers has helped a lot, but the settings that I changed made clearly made the biggest difference. Thanks for the feedback!

 

[frater secessus]

Sounds like you've got a great handle on it. And I fully support driving those batteries into the ground before buying new ones.

Thanks for the followup.

 

[jimindenver]

My only concern would be a hour of bulk and absorption is really short. Usually it takes time to do the bulk and then absorption is a set time. Mine are 3 hours unless the bank dips below a certain point.

The only time you should see a difference between charge times is when you use more power. A shallow discharge on a big bank is replaced by the same amount of power as the same discharge on a smaller bank.

 

[Canine]

jimindenver, I have my absorption time set at 5 hours, which is a little higher than the 4 hour recommendation. However, here is an interesting difference: I have the ProRemote (S) (a control panel that allows more detailed settings and provides more information and feedback). My understanding is the CC will automatically drop out of the absorption stage and into float once it reaches a certain state of charge. If all it takes is an hour to reach a specific charge, then it ignores the 5 hour setting then jumps into float mode. This option is labeled "Float Transition Current". If I set the Float Transition Current to "0" then it disables that function and will default to the manual 5 hour absorption stage setting. The absorption stage will then be a specific time like you stated.

When summer draws near, I will drop my absorption voltage of 14.4 down to 14.2 or lower. If the batteries don't need that high of voltage to be fully charged every day, I may as well change it at a more friendly voltage. The absorption recommendation for this battery bank is 13.9-14.4 volts. I tried the 13.9 and the bank wouldn't fully charge on a sunny day. Almost, but not quite. During the winter it's hard to get enough sun; therefore, the higher voltage settings to push in more juice. Works very well. Yes, that is a small bit harder on the batteries, but that's the price I pay for being in an environment with little sun. I don't bring my batteries below 30% discharge and the batteries aren't in a hot environment, so I'm not concerned about overheating. If I didn't have propane backup, my batteries would be toast. I really wanted to be 100% solar in the winter, but it's quite difficult. If the days stayed sunny, I could do it, but those completely overcast days kick my butt. Today was overcast, and output was only 2 amps. Seemed low, so I looked at my panels. One had goose poop on it and they were a bit dirty. I cleaned them and that brought the amps up to 6. Not enough to kick it out of bulk mode, but still getting a decent charge. Tomorrow is partly sunny and probably will go into float.

 

[John61CT]

I have the ProRemote (S) (a control panel that allows more detailed settings and provides more information and feedback). My understanding is the CC will automatically drop out of the absorption stage and into float once it reaches a certain state of charge. If all it takes is an hour to reach a specific charge, then it ignores the 5 hour setting then jumps into float mode. This option is labeled "Float Transition Current". If I set the Float Transition Current to "0" then it disables that function and will default to the manual 5 hour absorption stage setting. The absorption stage will then be a specific time like you stated.

aka using endAmps to end Absorb, by far the best way

But only works, assuming loads are present

if it is measuring that current via a shunt at the bank

not referencing current output by the SC.
In other words you have to base it on the proportion og current output that is actually being absorbed by the bank.

Is that the case with yours?

Does it also try to estimate bank SoC (state of charge) as energy is drawn down by loads over night?

 

[Canine]

John61CT, I don't know specifically how the the CC estimates the energy the batteries need. What I can say is it uses an algebraic formula to calculate the needs. Yes, it has a shunt. Without the shunt it would not be able to estimate what energy is used and what is needed. It has a "fuel gauge" in the form of a bar graph and displays a digital number that shows how many amps has been used and another digital number that shows how many amps it needs to be full again. It is pretty accurate. As I use energy during the night, the control panel will show how many amps are used and how many amps away from full the battery bank is. The bar graph (fuel gauge) also goes down from 100% to 99%, 98% and so on. The CC doesn't know how much the battery needs until it run a few cycles. It learns as it goes. As the battery begins to weaken, the "fuel gauge" adjusts; the older the batteries get, the smaller the "fuel tank". I hope this answers your question.

 

[Weight]

Here is another view, from experience, about off grid charging.;; https://handybobsolar.wordpress.com/2018-warning-turn-the-voltage-down/

 

[Canine]

Weight, I really like his site. He knows a lot. On my CC I have already turned down the max volts the batteries can receive to 14.4 so the bulk mode can't be any higher than that. No matter what changes I now make or what the temperature sensor decides to do, the max is 14.4.

 

[John61CT]

OK found and read the docs.

Nice gear!

Besides endAmps and profile customization, the IPN-Remote allows load cutoff to be based on bank AH SoC rather than just voltage drop.

I didn't see energy coefficient nor Peukert exponent settings for the battery monitor functions, so don't count on that being too accurate, even the most accurate AH-counting SoC monitors are often off by 5-10% between "Full reset" cycles.

Also, you need to estimate actual AH capacity as that declines over time, which unfortunately isn't going to be accurate without periodic 20-hour load testing.

Since you obviously don't mind spending the money on such issues, you may find it interesting to put a Merlin SmartGauge on the bank as a point of comparison. It is the most accurate SoC monitor on the market and has none of the above complexities of the coulomb-counting types.

Marketed and supported in NA by Balmar, but available cheaper ordered from the UK, depending on the exchange rate and shipping of course.

For a lot more info on these topics, search for the term "battery monitor" here https://marinehowto.com , three relevant articles there.

Everything about DC electrics on that site is well worth reading.

He (Maine Sail, CMS) also happens to sell the SmartGauge, although it's not the cheapest price, it's great to help support his amazing efforts educating the community.

 

[Canine]

John61CT, the fuel gauge will show 100% when it gets close to going into float mode, so it is not 100% accurate. I like to see it sit in float mode for a while to know I'm getting a true 100% fill. It is still useful; if it says 75% then I know I'm getting close to my limit. However, looking at the volts works well, too. If voltmeter says 12.4 volts, which is the same as 75% on the fuel gauge, then I still know I'm getting close to my limit. The voltmeter is a good way to tell as well, but the fuel gauge is more nuanced. I don't freak out if I hit 69% discharge. As long is I don't repeatedly go well below 70%, I'm happy.

I bought all this stuff when I had the money. Now I don't so I won't be getting the SmartGauge. Would be a fun tool, though.

Is it really necessary to do a load test over a 20 hour period to get an accurate result? Can a load test (such as day-to-day usage) over a shorter period of time be calculated then that information extrapolated for a 20 hour time period and still be accurate? That must be what my CC does. That seems like a lot to ask while still being accurate, but there are some smart people out there. However, my batteries never get a full rest. The fridge, the fan, or the heater is going on and off at any time all day and night. Surely, that must make calculating an accurate battery capacity more difficult.

 

[John61CT]

I like to see it sit in float mode for a while to know I'm getting a true 100% fill.

For optimum longevity, all your charge sources should be calibrated so nothing drops to Float voltage until *after* the bank is 100% Full.

Never mind what your gear says, your batt mfg should spec the endAmps definition of that Full point. Examples include "trailing amps stops falling, no chages for at least an hour" or maybe falling to .05C, so 4A for an 800AH bank.
Once volts drop to Float there's no more charging going on, just carrying consumer loads. With solar this point is "free energy" load dump time, heating water, making ice cubes, crank the stereo

For those whose charge sources are egg-timer based, they need to keep recalibrating based on their ammeter to make sure Absorb hold time stays right as their usage patterns change.

Having the SC under control of a BM is much better. Bogart Trimetric and Whizbang Jr are the only other OTS ones I know of.

No shore chargers do that afaik.

However, looking at the volts works well, too. If voltmeter says 12.4 volts, which is the same as 75% on the fuel gauge, then I still know I'm getting close to my limit. The voltmeter is a good way to tell as well, but the fuel gauge is more nuanced. I don't freak out if I hit 69% discharge. As long is I don't repeatedly go well below 70%, I'm happy.

50% is industry standard, going a little below even that is no cause for stress IMO as long as it's rare.

But volts alone won't do, unless usage patterns are very consistent, the owner has a proper SoC meter to get to know the relationships over time, otherwise you're flying blind.

And things do change radically as SoH declines.

Is it really necessary to do a load test over a 20 hour period to get an accurate result? Can a load test (such as day-to-day usage) over a shorter period of time be calculated then that information extrapolated for a 20 hour time period and still be accurate? That must be what my CC does.

I doubt it. Are you saying there is no place for you to input AH capacity?

SmartGauge is the only accurate SoC meter that automatically adjusts AH capacity over time, and calibrating an AH based unit against that would allow for some guesstimation.

But no, the only accurate measure of current AH capacity is a standard 20-hour load test.

Should be done at commissioning time, as well as 2-3 months after, with a proper breaking in protocol, capacity will usually increase a fair bit from their state upon delivery.

I suppose if a shorter load test is also done at those times, at least you'd have a standard to benchmark against, but it would be less accurate.

 

[John61CT]

However, my batteries never get a full rest. The fridge, the fan, or the heater is going on and off at any time all day and night. Surely, that must make calculating an accurate battery capacity more difficult.

Well the bank has to be offline for a proper load test, but then you're on shore power anyway.

SmartGauge does lose a little accuracy, up to 4% when loads are active

During charging maybe 8%

but no long resting period is required to correct for that, just isolate for a few minutes.

And that's as accurate as these things get, until you get into lab-class gear costing many thousands.

 

[Canine]

John61CT, yes I can input amp hour capacity. Factory default was 220 and I changed it to 420. Since that is a set number with no auto adjust option, that seems to be a static number. The manual, however, does specify that it learns over time and becomes more accurate with more use. How that works is unknown to me. How accurate that claim is is also an unknown. Blue Sky Energy does make some good stuff, and they seem to be an integral company.

If my "fuel gauge" starts getting really wonky, would I still be able to default to the voltmeter reading on the CC to get an accurate feel for how full the battery is? I understand how the volts read when there is a load. A battery that is weak can read 12.7 at rest, but under load quickly fall to 11.9 (and less) if it is weak- a sign of a bad battery.

I just found a good page on how a 20-hour load test is done. Doing an accurate 20-hour load test is quite involved and time consuming; I see why you mentioned needing to be on shore power during testing. I assume that is what you meant as I am totally off grid.

While the CC may lose accuracy over time, I wonder how much the accumulation of inaccuracy will be a problem. If I'm running my microwave and heating water without difficulty for two or three days of partly cloudy weather and the meter staying roughly accurate, then that's pretty good for my expectations. The most accurate meter will be when my inverter starts beeping because the battery bank can no longer keep up with microwave.

My biggest immediate concern is getting the appropriate amount of power to the batteries to keep them happy so they last as long as possible. Time will tell on that one. I haven't fried them yet and can still put big loads on there with no worries, so that is a good sign. Investing in good equipment and trying to squeeze as much solar as I can out of the system will pay for itself in battery longevity if I do my part correctly. That's why I went the extra mile to get two CCs, 570 watts of solar, and such. I did cut corners on the batteries, but that was a calculated decision. They are holding up to my expectations (despite my mistakes so far), but if I end up giving them a short life, I figure those are the ones I learned on. That has value. Then I can properly take care of the next set. I really like to take care of my equipment and use it the best way possible within what I can afford. Sometimes that takes a some effort, but I usually get there.

 

[John61CT]

Blue Sky Energy does make some good stuff, and they seem to be an integral company.

If my "fuel gauge" starts getting really wonky, would I still be able to default to the voltmeter reading on the CC to get an accurate feel for how full the battery is?

The market forces some marketing "leeway" doesn't mean they're lying, but I would update the AH input after 18-24 months assuming the bank's truly coddled, then yearly, after 4-5 years every six months.

The voltage vs SoC relationship will slowly change over time, or with unusual current levels.

I just found a good page on how a 20-hour load test is done.

Best I found is Maine Sail's

Search for 20 hour capacity test here http://www.pbase.com/mainecruising/programming_a_battery_monitor

Not sure if that made it to his new site

While the CC may lose accuracy over time, I wonder how much the accumulation of inaccuracy will be a problem.

You should have the ability to manually tell the BM 100% Full when you have confirmed mfg endAmps spec is reached.

Try to do that a couple times per week, both for bank longevity and maintaining BM accuracy.

The charge efficiency and Peukert exponent not taken into account, you can't do anything about that.

The residual AH capacity if you can't load test guesstimate, too low better than too high.

Or save for a SmartGauge over time to calibrate against, I sometimes see them for $120-140.

The most accurate meter will be when my inverter starts beeping because the battery bank can no longer keep up with microwave.

That's a failsafe LVD but just based on voltage, which is not SoC.

I did cut corners on the batteries, but that was a calculated decision.

Yes first bank should be cheap, don't invest in top-notch until you really know you know what you're doing.

Replace when SoH gets down to 70-75% to avoid "unexpected sudden failure".

Even by that standard a top bank if coddled can last 8-10 years.

 

[Canine]

The inverter beeping definitely would be after the batteries had received a full SoC, yet it still beeped. I will look into adjusting the AH input periodically. Not having to be so careful with usage has made the whole system more enjoyable. Having lots of solar wattage and being able to bring it in efficiently has been a great choice. Makes keeping the batteries happy much easier.

Thanks for the conversation everyone! I learned a lot. Now I'm going to finish absorbing what I learned

 

[John61CT]

If the bank is (apparently) full and your usual loads still pull voltage down to ?what? 11.something

then more AH bigger and/or better (LFP?) batteries are required

Or maybe the beeping means the AC load is trying to pull a higher *current* than the inverter wants to handle?