Off Grid Solar Upgrade

[paulca]

Time has come to upgrade my 50W panel.  The plan was to buy a rail kit for the garage roof and put 2 100W Renogy panels up there, with the option of adding a 3rd.

Then I noticed the 2x100W renogy panels is the same price as large 36V 300+W panel.

I knew ahead of time my 10A MPPT controller will not handle either on 12V.  10A limits it to 120W@12V charge current.  Call it 144W on boost.

I was intending on upgrading it to the 40A version... which then has a knock on effect to the battery pack.  A 12V Lead Acid will not enjoy 30A charge current.  A lithium pack would.

So it comes down to a choice between

- a single 100Ah 12V Lithium pack £440-£500.
- 4x 100Ah lead acids in 2P2S 24V config.  £400-£500.

The Lithium will take the 30A charge current from the 330W panel fine and will provide a usable 1200Wh give or take.

The lead acid however, at 24V will only need to take 15A per parallel pair, giving it about a 7.5A charge per battery.  However, because of lead acids depth of discharge limitations, it too would only provide 1200Wh of usable power.

I figure that lithium wins, because the Lead Acid on daily cycles 365 times a year will probably be losing 10-20% per year and will need replaced every 3 years or so.  The Lithiums LiFePO4 are claiming 3000-4000 cycles.  Giving them a life more like 8-10years.  (details of those stats aside).

[m k]

I have no experience how lithium stuff goes with their deep cycles but lead acid cases are not promising, or brochures were optimistic.
So it's Lithium from that but 100 vs 400.

If even storage then 1/4 price for lead acid and so a bit longer with the money.

[Renate]

100W solar panels are a joke. 300W solar panels are a commodity item. If you can fit a 300W, use that.

I currently run 2 x 300W into 2 x 6V 224 AH AGM batteries. If I were to build this today, I would go with LiFePO4. But this runs just fine, I have enough power. In a year or two I'll replace with LiFePO4.

The big advantage to my mind is that LA batteries only take a charge grudgingly. It's not a case of how high a current they can take, it's a case of how much current they will take. Let's say that your batteries are discharged and it's been raining all morning and the sun finally comes out at 2 PM. A LiFePO4 will say, "Let's make hay while the sun shines, give me all the current that you've got". A LA battery will start to feel full pretty quickly and will only want to take 5-10 Amperes which it will want to sip on for hours. You don't have hours of daylight left.

[m k]

Are there any serial to parallel and back systems for recharging?
Like 6V and 100A for recharge and then back to 60V and 10A of normal operation.

[paulca]

100W solar panels are a joke. 300W solar panels are a commodity item. If you can fit a 300W, use that.

I currently run 2 x 300W into 2 x 6V 224 AH AGM batteries. If I were to build this today, I would go with LiFePO4. But this runs just fine, I have enough power. In a year or two I'll replace with LiFePO4.

The big advantage to my mind is that LA batteries only take a charge grudgingly. It's not a case of how high a current they can take, it's a case of how much current they will take. Let's say that your batteries are discharged and it's been raining all morning and the sun finally comes out at 2 PM. A LiFePO4 will say, "Let's make hay while the sun shines, give me all the current that you've got". A LA battery will start to feel full pretty quickly and will only want to take 5-10 Amperes which it will want to sip on for hours. You don't have hours of daylight left.

If I put the 300W panel on my present setup it would work fine, but, the charge controller would very quickly ramp the panel voltage up towards open circuit to keep the current at 10Amps.  That's fine though, a 36V 300W panel won't make more than 10 amps.  However the charge side current limiter will override the MPPT and raise the panel voltage even further to only have 10Amps at 14.4V.  Basically 144W is my maximum through put, so the panel will spend most of it's time way out of the power band when in direct sun light.

But it does mean I can do things in steps, rather than all at once.

[Renate]

Well, yeah, but you're way up north and your panel is not physically tracking the sun.
300 Watts out of a 300 Watt panel is noon in the Saharan desert.
I hear that it's sometimes even cloudy in the UK?

I don't understand the concern for panel voltage or "power band".
Your charge controller and the voltage into the battery is what's important

Here is a 24 hour plot of my system. It's not yet spring, of course. A plot like this really shows how lonely the sunshine is.
You can see that I've used about 2/3 of the available sun power. (The little orange dots are what I drew in for a normal maximum sun curve.)
You can see at 11:15 my batteries didn't feel like taking any more charge. The voltage into the (LA) batteries really doesn't determine the current as much as you'd think.
When the converter goes from 14.4 to 13.8 at 14:20 the current hardly drops. at 14:30, I shut off the 125 Watt load (my PC).
At 15:30 I was using the PC on and off and the clouds were coming in.

[Ian.M]

Yep.  If Paulca's flag indicates his location, he's in the UK, known for having 'weather' rather then 'climate'.  If you can afford it, and the regulator/charge controller can handle the current, oversizing your panels is just common sense due to the high proportion of days without sustained direct sunlight, or with the sunlight compromised by Altostratus.

[tautech]

Jumping in to add a little personal experience where I've been closely involved with the design and construction of 2 remote data relay sites, one only solar and the other wind and solar for which a combined solution although it's quite exposed wasn't the ultimate solution and instead we ditched the turbine and took advantage the panel mount was future proofed to be able to add a 2nd panel.
This smaller installation used 2x 195AH SLA's paralleled for 12V predominately to make use of the 12V wind turbine however we reconfigured it to a 24V system and went to only solar(2x 300W panels) and although we used a suboptimal east facing installation with clear LOS to the horizon this has proved excellent for battery recovery early in the day with peak charges approaching 20A not long after sunrise.
Both installations have a quite steady load 24/7 so the earlier the batteries a brought to a float charge state the better as even without direct sunlight on the panels late in the day there's still sufficient panel power to manage the quiescent load from the system.

The larger installation has 3x 300W panels and 4x 165AH SLA's in a series/parallel configuration for 24V however like a spiders web this installation has grown with a higher HW load on it than was the original design and even with an optimized panel orientation successive poor charge days can impact dramatically on reserve capacity where we may have to run a generator on the 4th day.
Typical daily charge is ~60AH but a couple of days of just half that may require 100+AH to get the batteries back to a float state and I've recorded 30A charge peaks on this setup.
Both installations use Morningstar Tristar 60 MPPT controllers with online data and we can monitor panel and battery voltages and charge in real time or examine 100 days of logging.
This setup has been running a few tears now and logging tells us we've captured 1MW/Hrs of charge.

What I've learnt is to overrate the solar probably 30% beyond what you think and so prevent high battery DOD to get the best service life from them. The batteries in the larger installation are hovering ~10% DOD so we don't expect a long service life from them yet logging hasn't shown they're suffering unduly yet.

A couple of pics attached FYI

[paulca]

I think there are several distinct aspects to this.

1.  Panel size versus charge controller versus battery sizes.
2.  Full system load cycle sizing.

For the later, as I live at 55*N in a country were we have about 150 words for rain and cloud.... I tend to go 10x load for the panels, ie, expecting 10% output averaged over 24 hours and probably 2 weeks storage if it's going to make winter.  I do see short term road signs with 100W panels, but they have a trailer full of 12V lead acids, last one I got a peeking to had 8 100Ah batteries.

In this case, there is no fixed load.  I would like to slowly progress to running a lot more DC stuff off it.  I also hope to add a battery charger on a smart plug to kick in if the battery gets low, so that the load is un-interrupted, when effectively switching back to mains.

It is the former "Panel size" versus what I have for a charge controller.  It is physically fused at 10Amps.  I plan a 40A one for later.

I am taking the point, I think?, while for an hour either side of noon on a warm sunny day it might be significantly limited at 10A when it could produce 30A, but... later when the sun or off angle or there is thin cloud the 300W panel will still be able to produce those 10A for much more of the day than say a 150W panel, which might hit the 10A at noon, but barely make 5A outside of that.

The plan is to go with a rail setup for 3x300W panels.  Initially install 1.  Once that is out of the way, I can choose if the budget goes on new batteries or new charge controller.

Longer term, if it gets to storing a usable few Kwh, I might put an inverter on it and run a few office plugs off it.

[tautech]

Careful selection of the utilities/HW supplied by your installation is important to minimize losses where in the data relay installations the only bit of gear doing all the work is a DC powered POE 6 port switch that before we upgraded to the licensed 11 GHz backhaul dish providing 1 GB link capability all other HW runs on 24V POE whereas the 900mm bigger dish needs 1A 48V POE which is the only invertor in the whole system.
It's installation immediately impacted on reserve capacity.

However this one installation is the backbone of the system with 4 active links to other high sites, some of which push data to other high sites 20km away.....I did mention growing cobwebs earlier.