Can someone have a look at my calculator and tell me where I've gone wrong?

[JustinW]

Hi all,

I'm reasonably confident that some of this is right but I'm not sure if I have applied efficiencies in the right places, I was wondering if someone that is pretty good on both excel and their solar calcs can have a look at this for me.

It's a small system:
4 x 160A/H 12V Batteries
A 100|20 Victron MPPT
A 370W REC solar panel

It has a 30W load, constant calculated duty cycle over 24H but I want to know how far I can push it. I've attempted to make a calculator that gives me some feedback but I think I am tripping up over whether or not I should be applying the battery efficiency when I am recharging.

As in my total Efficient Battery W/h capacity with DOD is 3916.8

Should I just be applying the raw load / recharge rate to it (vastly oversimplified)
Or should I be applying the efficiency again for both removal and replacement of energy to my batteries? i.e. applying a multiplication factor of .85 to my daily MPPT output to determine what it can actually put back in?

I've attached my calculator if anyone cares to look. Hopefully it makes more sense then, it doesn't to me anymore, I've been looking at it too long...
Light blue cells are for data entry, light yellow are calculated

I do understand that this is a whole specialized field and I've probably missed a ton of concepts I don't even know I don't know but knowing I'm on the right track or how to get on the right track would be great!

TIA

Justin

 

[Supervstech]

Sorry, the spreadsheet is too complicated for me.
What are you trying to understand?

 

[Supervstech]

I dont understand battery efficiency.

There is inverter efficiency

There is pv conversion efficiency...

 

[Supervstech]

30W load, is that the standby drain of the inverter?

 

[Supervstech]

What voltage is your battery bank?

 

[wme]

Is there battery efficiency? What I have seen is energy density, KWH per kilo

 

[JustinW]

Hey, thanks for the replies.
From what I have read, there are efficiencies tied up with batteries, they're not always stated on datasheets etc. Sometimes they are.
I've found information scattered about the internet in regard to the efficiency of AGM batteries and it is typically understood to be 85 to 90%.
I read this as every 10W/h I apply to my battery, somewhere in the region of 8.5w to 9w is absorbed. Same with the retrieval of energy. For every 10W I want, I actually get 8.5 to 9 usable W/h.
The other 10 to 15% is converted to heat or such and unable to be reabsorbed by recollection systems within the battery.

For example, if I have 10000W/h of raw storage, disregarding DoD etc, whatever battery arrangement.
I actually have 8500w/H of usable storage.

Do I apply that effiency once at the start and just use raw numbers for everything after, or should I be applying that .85 every time I'm calculating a retrieval or storage process.
Or not applying it at the start, using to 10000W/h and just applying it to storage and retrieval.

I think the number is usually tied up in arbitrary factors that are not always well explained.

I've had a little more of a hunt around and have found this site that provides many good explanations far better than I can:

Instructions | PVEducation

www.pveducation.org

So after much more reading and revisiting of information I've settled on applying the efficiency for retrieval and storage operations only.

 

[MichaelK]

Your spreadsheet was mostly a waste of time. Here is where your biggest failure is...


Assuming you get 8.8 hours of sunlight to put into your sheet is totally wrong. Just because the sun is up for 8.8 hours does NOT mean your panels are are full production for 8.8hr.

Research into the concept of the sunhour (sh), and find out what the numbers are for your area. In Perth, you are about as far South as I am above north. I'd suggest that in June, you are likely to get only 3.0sh, and maybe 6.0sh in December, if your panel is pointed due North. So your production calculations greatly exceed what you are likely to make in the real-world.

I'd say the 75% efficiency number is about right, though I typically apply 85% myself. So, in the real-world, your daily production is more likely to be 1100Wh in June as compared to the 2442Wh you calculate.

A second problem for you is the ratio of panel to battery capacity, which is also way off. Are these batteries lead-acid or Li? Using a more conservative 1/8th of C value for lead-acid, and utilizing an 85% fudgefactor for panel efficiency, what you really should have is....
[(160Ah X 4 batteries)/8] X 12.5Vcharging = 1000W. With the fudgefactor, that's actually 1176W.

So, in reality, you need three of those panels to meet that proper requirements for that battery bank. You'll need to get rid of that little 20A charge controller too. Get something in the 75-100A range.

So, to sum it all up, your output calculations are off by ~100%, and your storage feeding rate is off by ~300%

 

[Supervstech]

Solar is NOT something to go into optimistically.

Realistic if not depressingly pessimistic yield far happier results.

Over estimate your loads, underestimate your production abilities, and double the safety factors…

 

[JustinW]

Your spreadsheet was mostly a waste of time. Here is where your biggest failure is...
View attachment 175379

Assuming you get 8.8 hours of sunlight to put into your sheet is totally wrong. Just because the sun is up for 8.8 hours does NOT mean your panels are are full production for 8.8hr.

Research into the concept of the sunhour (sh), and find out what the numbers are for your area. In Perth, you are about as far South as I am above north. I'd suggest that in June, you are likely to get only 3.0sh, and maybe 6.0sh in December, if your panel is pointed due North. So your production calculations greatly exceed what you are likely to make in the real-world.

I'd say the 75% efficiency number is about right, though I typically apply 85% myself. So, in the real-world, your daily production is more likely to be 1100Wh in June as compared to the 2442Wh you calculate.

A second problem for you is the ratio of panel to battery capacity, which is also way off. Are these batteries lead-acid or Li? Using a more conservative 1/8th of C value for lead-acid, and utilizing an 85% fudgefactor for panel efficiency, what you really should have is....
[(160Ah X 4 batteries)/8] X 12.5Vcharging = 1000W. With the fudgefactor, that's actually 1176W.

So, in reality, you need three of those panels to meet that proper requirements for that battery bank. You'll need to get rid of that little 20A charge controller too. Get something in the 75-100A range.

So, to sum it all up, your output calculations are off by ~100%, and your storage feeding rate is off by ~300%

Thanks, this is helpful.

I advanced it a lot further yesterday and consulted a couple of other experts about insolation and have made some improvements, mainly vastly reducing peak solar radiation hours.
It is extremely sunny where it is, Newman Western Australia, so they see min 6h ish in winter.
I've also taken into account peukert constants for the batteries, they're AGM.
I also split the day night load apart.
E.g. day load has the mppt inefficiency applied to it, then the remained + battery inefficiency is returned to the battery.
Night load is direct load + inefficiency and total charge/discharge rates are calculated from that.
I can DM the finished product to you if you are interested

 

[MichaelK]

It is extremely sunny where it is, Newman Western Australia, so they see min 6h ish in winter.

I don't think you understand what I am talking about. Southern California where I'm located is also extremely sunny.

It is not about how many hours of daylight you get. It's about the angle of the sun hitting the panels. Here is what is happening. Let's assume your panels are facing due North.

At 8am, when the sun is low to the East, the sunlight hitting your panels is at a very acute angle, almost 90 degrees. At 8am, very little of the light is hitting your panels, so they are only putting out 5-10% of their rated output. At 9am, the sun is a little further North, and maybe now they are at about 25% output. At 10am, maybe 50%, 11am 75%, and noon 90+%.

If you had 1000W of panels, what they would be putting out at each hour would be ~
8am 100W
9am 250W
10am 500W
11am 750W
12noon 900W
1pm 750W
2pm 500W
3pm 250W
4pm 100W

The total in this example is 4100Wh. Divide the 4100Wh of power, by the 1000W of panels, and you get 4.1sunhours.

Looking up Perth, I see you are getting a yearly average of 5.7sh. Winter is lower, summer is higher. Actually slightly better than I predicted.

 

[Hedges]

4 x 160A/H 12V Batteries
A 100|20 Victron MPPT
A 370W REC solar panel

A second problem for you is the ratio of panel to battery capacity, which is also way off. Are these batteries lead-acid or Li? Using a more conservative 1/8th of C value for lead-acid, and utilizing an 85% fudgefactor for panel efficiency, what you really should have is....
[(160Ah X 4 batteries)/8] X 12.5Vcharging = 1000W. With the fudgefactor, that's actually 1176W.

What charge rate does the manual say your batteries want?
My SunXtender prefer at least 0.2C, which would be 1500W for your size bank. Around 100A or so charger as was mentioned.
Since PV panels might make 75% or 85%, that's 2000W of panels.
However, they'll only make that for a couple hours middle of the day. Want to charge for about 4 hours, then absorb with lower current for a couple more hours.
About 3000W of panels, 1500W tilted East and 1500W tilted West, would do nicely. If you aren't drawing much power during the day; add more to cover that.

Battery efficiency says you have to put in more watt hours than you consumed.
It doesn't further reduce how much you can draw from a fully charged battery, beyond the DoD you select.
Peukert would reduce how many watt hours you could draw at high amperage.

Although if you really have just 30W load, 720 W, then you could just keep them at float. Your single 370W panel could be sufficient with at least 3 effective sun hours.
You could run a week with no sun, although would rather recharge in a day.

 

[JustinW]

I understand it's not about the daylight hours.

These are the profiles for site, it's one of the sunniest places on earth.


They're much closer to here than Perth


I've identified the upper limit as 60W and anything higher will give me a run flat in the dead of winter before the monthly average picks up again. I wouldn't run it close to that, 50W max is going to the person that's trying to over-leverage this setup with explicit warnings that extended cloud cover or unseasonable weather will see increased risk of failure

It should hopefully outperform 50W though as any I haven't accounted for any bonus efficiencies gained time over the 5.72h
As the static load will be able take power in some form from the solar for 10-11 hours of the day instead of having to take it all from the battery, which is more inefficient.

Charge rate is .2C preferred but with the setup it will never reach it.

It's skid mounted, not premises mounted and my footprint is set and I can't up size my panels at all. I could probably gain a lot from a LiPO4 setup but for the extra 10K I think I'll pass and tell the people trying to over leverage it to stop doing it.