Pop Quiz

[AgroVenturesPeru]

I'd like to know what the best panel array would be in my situation (let's disregard the economic aspects of the decision). I'm concerned in knowing which option is the most efficient. I think I know the answer, but want to know what people with more experience would select. Also, feel free to suggest your own alternative idea, if you feel there is a better option.

This location is 6 degrees from the equator and 3,000feet elevation, but cloudy days are common.
This place will probably never see a temperature below 40 degrees Fahrenheit.


Needs: To charge a 48v, 9.6kwh battery bank (these are the useable kwh). Total kwh of battery = 10.5kwh

Here is a picture of the MPPT/Inverter confines I need to work within:



A)
12; 315W Monocrystalline Panels (3,780W total)
Arrangement: 3s4p (s=series; p=parallel); Totals: 101.25V / 37.36A
41.18Voc
33.75Vmp
9.99A (Isc)
9.34A (Imp)

B)
8; 450W Monocrystalline Panels (3,600W total)
Arrangement: 2s4p; Totals: 81.3V / 44.28A
49.57Voc
40.65Vmp
11.49A (Isc)
11.07A (Imp)

C)
9; 400W Monocrystalline Panels (3,600W total)
Arrangement: 3s3p; Totals: 121.8V / 29.58A
48.64Voc
40.6Vmp
10.3A (Isc)
9.86 (Imp)

D)
12; 320W Polycrystalline Panels (3,840W total)
Arrangement: 3s4p; Totals: 94.68V / 32.28A
36.62Voc
31.56Vmp
8.88A (Isc)
8.07A (Imp)


Bonus question: How could the charge controller have an allowance for the maximum PV array power of 4500W when the maximum solar charge current is only 80A?

 

[snoobler]

You're focusing on incredibly tiny details. The four options are efficient within 1% of each other. Efficiency diferences between all of them are negligible with the higher Vmp options being SLIGHTLY less efficient than the lower ones...SLIGHTLY. And this could be offset by improved voltage drop in the PV wiring to the controller.

Even with 40°F, you'll need a 10% margin on your Voc, i.e., 145/1.1 = 132V

Looks like all of them comply with that.

D) is likely the best option since you have the most power.

Bonus: 80A * 56.25V (actual charge votlage, not system voltage) = 4,500W, but you will only pull that at the higher voltages.

If you just look at 80A * 48V = 3,840W

On a sunny day, you should be able to charge the battery 1.5X

 

[AgroVenturesPeru]

You're focusing on incredibly tiny details. The four options are efficient within 1% of each other. Efficiency diferences between all of them are negligible with the higher Vmp options being SLIGHTLY less efficient than the lower ones...SLIGHTLY. And this could be offset by improved voltage drop in the PV wiring to the controller.

Even with 40°F, you'll need a 10% margin on your Voc, i.e., 145/1.1 = 132V

Looks like all of them comply with that.

D) is likely the best option since you have the most power.

Bonus: 80A * 56.25V (actual charge votlage, not system voltage) = 4,500W, but you will only pull that at the higher voltages.

If you just look at 80A * 48V = 3,840W

On a sunny day, you should be able to charge the battery 1.5X

Interesting response. And it's not what I was thinking. As you noted: 80A x 48V = 3,840W. This is the maximum I could supposedly produce before exceeding the amperage output limit of the charger. That's why I decided against D. I figured it doesn't leave any breathing room.
Also, D only yields a working voltage and amperage of 94.68 and 32.28 respectively. 60-115VDC is the operating voltage range of the MPPT, and "D" doesn't seem to harness enough of the higher end of this range.

Personally I chose A, because of the 101.25V it produces. Perhaps my understanding is off. But I assumed it better to produce higher voltage from each series, because it would perform better on cloudy days.

B was not as productive in this regard.

It couldn't be C, because C produces 121.8V, which exceeds the operating voltage range of the MPPT.

I do appreciate your patience in looking at all the data points. I know it's tedious, but your feedback helps me check my comprehension. So thank you.

One other small detail: You mention the temperature might require 10% buffer for the VOC. Shouldn't you multiply by 1.1 instead of dividing? I.e. when the temperature gets colder the VOC increases.

 

[Hedges]

"best panel array would be"

Best in what sense, harvest the most power with the limitation of just one charge controller?
Add more strings in parallel. The charge controller will never put out more than 80A, just clip during the middle of the day, but it'll put out the full 80A longer.
And, I always point out that you can break your parallel strings into two or more groups.
If you tilt half East toward 10:00 AM sun, and half west toward 4:00 PM sun, they will be at 90 degrees to each other.
You can then put in 50% more panels for the same peak current. Power production will be more uniform throughout the day, for more total power.
You can also aim some more North and some more South, leveling production over the seasons.

Panels are cheap, batteries are expensive. If your draw is during the day, this will reduce cycling of batteries and extend their life.

 

[snoobler]

Interesting response. And it's not what I was thinking. As you noted: 80A x 48V = 3,840W. (1) This is the maximum I could supposedly produce before exceeding the amperage output limit of the charger. That's why I decided against D. I figured it doesn't leave any breathing room.
(2) Also, D only yields a working voltage and amperage of 94.68 and 32.28 respectively. 60-115VDC is the operating voltage range of the MPPT, and "D" doesn't seem to harness enough of the higher end of this range.

(3) Personally I chose A, because of the 101.25V it produces. Perhaps my understanding is off. But I assumed it better to produce higher voltage from each series, because it would perform better on cloudy days.

B was not as productive in this regard.

It couldn't be C, because C produces 121.8V, which exceeds the operating voltage range of the MPPT.

I do appreciate your patience in looking at all the data points. I know it's tedious, but your feedback helps me check my comprehension. So thank you.

(4) One other small detail: You mention the temperature might require 10% buffer for the VOC. Shouldn't you multiply by 1.1 instead of dividing? I.e. when the temperature gets colder the VOC increases.

(1) No. That's the maximum at nominal battery voltage. The SCC lists 4500W because it can deliver the 80A at HIGHER than 48V. You charge a 48V battery to much higher than 48V. The SCC can deliver 80A up to 56.25V, thus 80A * 56.25V = 4500W. Furthermore, SCC can accept more panels than they are rated for provided you don't exceed the input limits. They just don't use anything over their ratings.

For example, my Victron 250/100 can accept 250V of panels and deliver 100A of charge current; however, it also has a 70A INPUT limit. The charger is limited to 58V * 100A =5800W of production; however, I could put 250V * 70A = 17,500W of panels on it, and it would just ignore anything over the 100A charge limit.

(2) & (3) "higher end of the range" - this isn't a thing. It's actually the opposite. MPPT is most efficient around 50% higher than battery voltage. When you move the voltage higher, you start incurring greater losses in the DC-DC conversion, typically 1-2%; however, one may recover these losses due to the improved voltage drop characteristics from higher voltage/lower current.

(4) I was dividing the LIMIT of the SCC by 1.1 yielding a new panel voltage limit, i.e., panel Voc must not exceed 132V to allow for cold temperature voltage spikes.

(D) is the option that will produce the most power because you have the most panels. The voltage parameters are acceptable for the SCC and optimal MPPT operation for a 48V battery. Again, there may be 1% variation in the efficiencies of the options, but "D" will give you the most solar collection.

 

[AgroVenturesPeru]

"best panel array would be"

Best in what sense, harvest the most power with the limitation of just one charge controller?
Add more strings in parallel. The charge controller will never put out more than 80A, just clip during the middle of the day, but it'll put out the full 80A longer.
And, I always point out that you can break your parallel strings into two or more groups.
If you tilt half East toward 10:00 AM sun, and half west toward 4:00 PM sun, they will be at 90 degrees to each other.
You can then put in 50% more panels for the same peak current. Power production will be more uniform throughout the day, for more total power.
You can also aim some more North and some more South, leveling production over the seasons.

Panels are cheap, batteries are expensive. If your draw is during the day, this will reduce cycling of batteries and extend their life.

Yes. On an overcast day.

 

[AgroVenturesPeru]

Even with 40°F, you'll need a 10% margin on your Voc, i.e., 145/1.1 = 132V

Looks like all of them comply with that.

What are you looking at to determine that? Let's take C for example with 48.64VOC. If I have three of these panels in series it would be 145.92Voc, right? This exceeds your 132V recommendation.

 

[snoobler]

Correct. I didn't catch that the first time through. I was just looking at the total number you provided assuming you used Voc. C) will not work.

Temperature change in resources:

PV array temperature correction table (NEC 2017)

There are calculators like this one made by @upnorthandpersonal which help you calculate PV array voltage and power for low temperatures based on the specific specifications of your panels. These are great tools and will give more precise...

diysolarforum.com

 

[Hedges]

What are you looking at to determine that? Let's take C for example with 48.64VOC. If I have three of these panels in series it would be 145.92Voc, right? This exceeds your 132V recommendation.

I would do the math in the opposite direction, to match reality and avoid confusion.
It is not that the charge controller accepts lower Voc on a cold day (132V recommendation). Must never exceed 145V.
It is that the PV panel delivers higher voltage on a cold day.

Panel (C) is 48.64 Voc at 25C ambient.
40F, (assuming that is the record coldest temperature ever) would be 4.44C, call it 20C below ambient.
Temperature coefficient varies between panels, so need the spec for this panel to be sure. I'll use -0.4%/degree C
48.64 x (1 - (-0.004 x 20)) = 52.53
3 panels x 52.53 = 157.5V
So I wouldn't risk that, at least not without getting the correct temperature coefficient.

But you can select other charge controller that accept 150, 200, 250 Voc. Some also tolerate higher voltages by disconnecting, waiting for panels to warm up and drop in voltage.

Update: Oops, option (C) exceeds 145V even at ambient!

 

[AgroVenturesPeru]

(2) & (3) "higher end of the range" - this isn't a thing.

(4) I was dividing the LIMIT of the SCC by 1.1 yielding a new panel voltage limit, i.e., panel Voc must not exceed 132V to allow for cold temperature voltage spikes.

Here's what I'm talking about when I say "higher end of the range"

---

What is the "range" on the following victron unit below? Also, what is the allowed input? Is it the same as your 70A input on your 100A output MPPT??? I'm still a bit confused on this topic.

If I use the same panels from "D" in my original post, but I increase the array to a 3s5p setup. I'll have a production of 94.68v and 40.35A...So is that okay for both this Victron and the above Growatt? Just when I thought I had it figured out, your response made me realize I was on the wrong track.

To determine charge controller compatibility, the first thing to calculate then is the Open Circuit Voltage of the PV array, then the input voltage & amperage it provides to the charge controller, then the amp output of the charge controller?

 

[snoobler]

Why are you switching products? That's a different unit with different design parameters. It's essentially designed for high voltage input similar to what one sees on a grid tie installation with only 1-2 series strings.

Victron doesn't publish MPPT range. The only criteria is that the panel voltage must be +5V above battery voltage before it will try to charge.

Again, you are focused on incredibly tiny details. Make sure your panel parameters meet the input requirements and shoot for a Vmp around 50-100% above battery voltage. You can be confident you're within 1% of optimal efficiency.

 

[AgroVenturesPeru]

Correct. I didn't catch that the first time through. I was just looking at the total number you provided assuming you used Voc. C) will not work.

Temperature change in resources:

PV array temperature correction table (NEC 2017)

There are calculators like this one made by @upnorthandpersonal which help you calculate PV array voltage and power for low temperatures based on the specific specifications of your panels. These are great tools and will give more precise...

diysolarforum.com

Yes, I thought your 10% buffer was a bit excessive.
For example I'm using this calculator: https://www.explorist.life/solar-charge-controller-calculator/

I was actually wrong about the VOC of the 320W panels (the vendor doesn't have the right specs on their webpage). According to the manufacturers specs, the panels actually have a VOC of 46.73 (not 36.62). I was thinking about getting 15 of them (3s5p). Their VOC would be 140.19 in that case. According to that calculator when I plug in all the parameters it tells me 49.88064044444444 Volts per Panel & 149.6419213333333 Total Volts in Array.

Is that cutting it too close? If it's a particularly frigid night, couldn't I just get up and disconnect the PV array from the charge controller, and wait till later the next morning (after it's gotten a bit warmer) to plug it in again?

 

[snoobler]

You going to trust that YOU are the "backup?" Are you certain that you will be there every morning that there might be a problem?

Options:

1. Buy another panel, go 2S8P
2. drop a panel and go 2S7P.
3. Buy a 200/250/450V/600V charge controller and push your voltage as high as you like.

If I were you, I'd do #1

 

[Hedges]

Yes, I thought your 10% buffer was a bit excessive.
For example I'm using this calculator: https://www.explorist.life/solar-charge-controller-calculator/

I was actually wrong about the VOC of the 320W panels (the vendor doesn't have the right specs on their webpage). According to the manufacturers specs, the panels actually have a VOC of 46.73 (not 36.62). I was thinking about getting 15 of them (3s5p). Their VOC would be 140.19 in that case. According to that calculator when I plug in all the parameters it tells me 49.88064044444444 Volts per Panel & 149.6419213333333 Total Volts in Array.

Is that cutting it too close? If it's a particularly frigid night, couldn't I just get up and disconnect the PV array from the charge controller, and wait till later the next morning (after it's gotten a bit warmer) to plug it in again?

For the 150 V max charge controller, not too close, no worries.
For the 145 V max charge controller, it is too close.
You won't be awake checking the thermometer, so you won't notice until your batteries get run down because charge controller died.

If you parallel some strings facing SE and some facing SW as I've suggested, I think that'll pull Voc a bit lower. Don't rely on that to get under the limit, but it should give additional margin.

 

[AgroVenturesPeru]

Why are you switching products? That's a different unit with different design parameters. It's essentially designed for high voltage input similar to what one sees on a grid tie installation with only 1-2 series strings.

Victron doesn't publish MPPT range. The only criteria is that the panel voltage must be +5V above battery voltage before it will try to charge.

Again, you are focused on incredibly tiny details. Make sure your panel parameters meet the input requirements and shoot for a Vmp around 50-100% above battery voltage. You can be confident you're within 1% of optimal efficiency.

I'd like to start out with a good system, max it out, and not have to adjust anything for the lifetime of the products. I'm just showing the options I have to work with.
There's:
Victron EasySolar 48/5000
A combination of Victron Charge controller and the Victron Multiplus 48/5000/70-100 or 48/3000/35-50 (not sure if the latter would be enough, as I don't understand what "35-50" refers to; I guess I could probably get by with just 3000watts of inverter though.)
Growatt SPF 5000TL HVM
Growatt SPF 5000 ES

I'm not really sure what to go with. Thanks for clearing up the minimum voltage requirement for the victron equipment though. I didn't know it was about 5V above 48V.

The LiFePo4 batteries here available locally are all 48V. There's 2 options:
3 Narada 48NPFC100 (100ah)
3 Pylontech US3000 (74ah)
I'm leaning towards the Pylontechs as they come with 10 years of warranty (Narada only has a one year warranty) and there is just more wealth of information online regarding the Pylontechs.

 

[AgroVenturesPeru]

You going to trust that YOU are the "backup?" Are you certain that you will be there every morning that there might be a problem?

Options:

1. Buy another panel, go 2S8P
2. drop a panel and go 2S7P.
3. Buy a 200/250/450V/600V charge controller and push your voltage as high as you like.

If I were you, I'd do #1

Ok, but this gets back to my original concern about designing for cloudy days. Wouldn't it be better to have larger series? Wouldn't 3s be better than 2s in that regard?

 

[AgroVenturesPeru]

For the 150 V max charge controller, not too close, no worries.
For the 145 V max charge controller, it is too close.
You won't be awake checking the thermometer, so you won't notice until your batteries get run down because charge controller died.

If you parallel some strings facing SE and some facing SW as I've suggested, I think that'll pull Voc a bit lower. Don't rely on that to get under the limit, but it should give additional margin.

Actually it rarely gets below 50 degrees in our area, so we're not planning on having a heater. If it were to get down to 40F I would definitely notice, because I would not be sleeping.

 

[snoobler]

Please provide a reference where it says you need higher voltage for cloudy days.

The existence of light on the panels provides the voltage. The intensity of the light provides the current. The max power is when V * A = a maximum, and that's usually right around Vmp.

 

[AgroVenturesPeru]

Fwiw, Here's the specs for the 320W Polycrystalline Panels:

 

[AgroVenturesPeru]

Please provide a reference where it says you need higher voltage for cloudy days.

The existence of light on the panels provides the voltage. The intensity of the light provides the current. The max power is when V * A = a maximum, and that's usually right around Vmp.

This video at the 2min mark: