Off-grid solar diagram / ideas

[jonescg]

I've decided to start three threads on how the keen amongst us might set up a solar + storage system.
Pending further input from the brainstrust here, this would be a fairly sensible way to do it:

An off-grid setup:

Proposed off-grid solar setup

offgrid setup.png (31.04 KiB) Viewed 2408 times

The PV array is set up as three panels in series, and as many of these in parallel as the roof (or budget) will allow.
Any more than two strings in parallel will call for in-line fuses on each array.

The isolator is a standard bit of kit. The MPPT solar charge controller manages the charging of a 48 volt battery. A good example is the Victron MPPT 150/100.
It comes with a LFP setting which sets the peak charge voltage at 57 volts and stops charging after that.

The load contactor is driven by the battery management system. If a cell goes out of spec (usually below 2.5 volts) the contactor will open. It will remain open until the voltage returns to within spec. The inverter is a separate bit of equipment which takes battery voltage and turns it into 230 V AC for the home.

It is possible to by inverters which both manage charge control and the inverter, but one thing living off-grid in my youth taught me, was that lights and refrigerator are well served by being on a DC supply. At least when the inverter craps out you can still find your way to the fridge to grab a beer

An example of such an inverter would be the MPP Solar PIP5048 inverter. It also has a 240 V AC input for the grid should you have it. This is where you use the grid as a generator, or an emergency fall-back. Or just to run the aircon. The 5048 doesn't push power back onto the grid (no export).

Any tips, changes, recommendations - feel free to chime in!

[bigjsl]

I'm basically building this.

The main contactor on the battery may be replaced by a good software control system but I appreciate the value in a contactor driven by a BMS.

Each solar panel is actual a string of three.

Storage System 1.png (742.23 KiB) Viewed 2384 times

[jonescg]

If I'm not mistaken, the way this is set up, should the BMS decide to cut the contactor off it will remain off even when the sun comes out again. If the BMS is the last line of defence this might be fine.

I think the PIP5048 can be set to stop drawing power at a setpoint voltage, like 43 volts or something. However as soon as the sun comes out the MPPT unit will start charging again.

[T1 Terry]

I've decided to start three threads on how the keen amongst us might set up a solar + storage system.
Pending further input from the brainstrust here, this would be a fairly sensible way to do it:

An off-grid setup:

offgrid%20setup.png (31.04 KiB) Viewed 2361 times

The PV array is set up as three panels in series, and as many of these in parallel as the roof (or budget) will allow.
Any more than two strings in parallel will call for in-line fuses on each array.

The isolator is a standard bit of kit. The MPPT solar charge controller manages the charging of a 48 volt battery. A good example is the Victron MPPT 150/100.
It comes with a LFP setting which sets the peak charge voltage at 57 volts and stops charging after that.

The load contactor is driven by the battery management system. If a cell goes out of spec (usually below 2.5 volts) the contactor will open. It will remain open until the voltage returns to within spec. The inverter is a separate bit of equipment which takes battery voltage and turns it into 230 V AC for the home.

It is possible to by inverters which both manage charge control and the inverter, but one thing living off-grid in my youth taught me, was that lights and refrigerator are well served by being on a DC supply. At least when the inverter craps out you can still find your way to the fridge to grab a beer

An example of such an inverter would be the MPP Solar PIP5048 inverter. It also has a 240 V AC input for the grid should you have it. This is where you use the grid as a generator, or an emergency fall-back. Or just to run the aircon. The 5048 doesn't push power back onto the grid (no export).

Any tips, changes, recommendations - feel free to chime in!

A better solar set up would be to link the number of panels in series to add up to close to the max open circuit voltage of the MPPT controller or between the min and max PV MPPT voltage range when the Vmp is added together in the series string - this is not always the same number of panels so best to check both totals before committing to a particular arrangement. Once the number of panels in the series string is determined then add panels in parallel to each group until the max input power adds up to the particular controllers capacity at the chosen battery voltage.
Why? If a single panel in a series/parallel string looses output due to shadow or what ever, the output from all the panels in that string are lost because that strings voltage is now less than the other parallel strings voltage so any output is blocked until the voltages are again equal. By adding the panels in parallel and then connecting these groups in series it would have less effect is one panel is shaded or in fact if a line of shade affected panels across panels in each parallel group.
My drawing takes even longer than my typing so although a picture says a thousand words I can type a lot of words in the length of time it takes me to draw a picture If this doesn't make sense then I'll attempt to draw what I'm trying to say and see if that explains it better.

T1 Terry

[bigjsl]

I was contemplating a diode across the contactor which took me off to the world of high current diodes.

I'm going to wait and see what Batrium can do. It has multiple outputs and inputs and places for logic. I've got the software running in a VM on my Mac so I'm almost ready to play.

In another part of my life I did some work for a well known Australian battery company attempting to get their high tech battery talking to Inverters. One of the problems was that if the battery disconnects from the Inverter it requires human intervention to get it back online. For some inverters the battery IS the power supply and disconnecting it leaves the Inverter in a standby state. This leads to perverse solutions like a 36 volt power supply across the bus with a series diode to keep enough volts (and amps) present to keep everything running.

That same organisation has a version of their battery with two contactors each with a diode in opposite polarities. If both contactors are open no current flows, if either is closed the battery can either charge or discharge but only when both contactors are closed can the battery both charge and discharge.

The thing that I was looking for then was Inverters and Chargers with inhibit or enable inputs or loops. I struggled to find them then. For this project I was finding lots of dumbish solar charge controllers and pure sine inverters that have serial ports but little by way of inputs on eBay. Once I found the PIP I was excited then I found the discussion group here and was stoked. Then I discovered that the only authentic PIPs I could find are pf1.0 and don't work with Coulomb et al's lovely firmware. I found something on AliExpress that looks like a Chinese clone which may have been a dumb option but it did come with a parallel kit which I will recycle if push comes to shove.

A cell over or under voltage in this particular pack is going to be a big problem because it's going to be damn hard to replace a single cell for much less than the total purchase price of the pack. We're all probably hoping for someone to have bought batteries and then fail to work out how to use them!

[bigjsl]

Terry, I think you need to be careful about putting panels in parallel.

I have read a few articles on the need for blocking diodes to prevent shaded panels from sinking current from generating panels.

https://www.electronics-tutorials.ws/di ... iodes.html

I have bought 25 250 watt panels recently including one which has a burnt out set of interconnects between adjacent cells so it's definitely a failure mode. Before that I was just planning to parallel panels.

For our application keeping Vmax below 120 volts lets us side-step quite a bit of electrical wiring rules because it's Extra Low Voltage. In particular it can be installed by a non electrician.

It appears you can get MC4 in-line fuses with integral diodes. I have some coming from China. I plan to use them on each string of three panels as I parallel four strings into 3KW, twice. That means 8 diodes.

jsl

[T1 Terry]

Terry, I think you need to be careful about putting panels in parallel.

I have read a few articles on the need for blocking diodes to prevent shaded panels from sinking current from generating panels.

https://www.electronics-tutorials.ws/di ... iodes.html

I have bought 25 250 watt panels recently including one which has a burnt out set of interconnects between adjacent cells so it's definitely a failure mode. Before that I was just planning to parallel panels.

For our application keeping Vmax below 120 volts lets us side-step quite a bit of electrical wiring rules because it's Extra Low Voltage. In particular it can be installed by a non electrician.

It appears you can get MC4 in-line fuses with integral diodes. I have some coming from China. I plan to use them on each string of three panels as I parallel four strings into 3KW, twice. That means 8 diodes.

jsl

Hi John, this one is an extremely contentious issue among members here regarding the requirement for blocking diodes. I use blocking diodes on the semi flexible panels because they have different modules to the rigid panels and can not handle much back voltage at all, but I also only ever connect these panels in parallel and never in series for the same reason.
We have been using the locally made Tindo solar panels for a while now and not had any issues with parallel/series connection so possibly a bit to do with the quality of the modules they use. They are poly panels and not mono and seem to better suit the rather high panel temps we see in Australia where the mono panels seem to better suit the cooler European and US climate.
We are not far out of Adelaide up in Mannum so if you are interested in a visit drop us a PM and we will arrange a time etc

T1 Terry

[bigjsl]

Hi John, this one is an extremely contentious issue among members here regarding the requirement for blocking diodes.

We are not far out of Adelaide up in Mannum so if you are interested in a visit drop us a PM and we will arrange a time etc

Love a good contentious issue! Can't see it doing me any harm and I have no idea how crappy my second hand panels are.

It would be fun to pay you a visit in Mannum. I will follow that up

jsl

[T1 Terry]

We actually do this stuff as a business but mostly RV stuff like house boats, motorhomes and caravans. We have done a few house jobs as well, the latest is a total off grid mansion of a holiday house built to cope with 5 families over the summer holidays and always someone there all yr.
Still fine tuning that one but it is a reasonably big set up, 10kW of solar, 6 MPPT controllers with 1600w roughly of solar on each, 1200Ah @ 24v of Winston LYP cells and 2 x 5kva inverters in parallel. It does have a 10kva genset wired in for auto run with a quiet time between 5pm and 9am the next morning. Yet to see if it starts up in anger because the battery hasn't dropped below 60% so far even after a few days of rain and overcast weather. 4 of the controllers are rated at 1600w @24v so yet to see if they have issues in the summer sun with the slight over capacity, but I expect the panel heat will knock the capacity down a bit so all should be fine, the 7kW air con might tax it a bit though depending how hard they run it.

T1 Terry

[T1 Terry]

Not sure if this is in the right thread so if it's in the wrong place please move it.
These are few photos of the battery/inverter/MPPT controller set up we installed for the off grid house I mentioned in the last post

Powerpack LH view.jpg (120.45 KiB) Viewed 2104 times

Powerpack RH view 2.jpg (110.23 KiB) Viewed 2104 times

1200Ah @ 24v battery pack.jpg (158.22 KiB) Viewed 2104 times

T1 Terry

[offgridQLD]

Where are you getting your 200ah Winston cells from?

[T1 Terry]

Where are you getting your 200ah Winston cells from?

Anyone who can supply them at the moment, but direct from the factory sooner rather than later we are hoping, if they come to their senses regarding price rather than the crazy prices/quantities they are talking at the moment. We have narrowed down a range we carry now to 4 size/capacity cells and may yet reduce that to 3 sizes to simplify ordering and keep up rapid stock turn over to avoid carrying old stock.

T1 Terry

[antiscab]

reviving an old thread here, but what blocking diodes do you guys use?

on my setup I have 4s5p 175W 36Vmp, 5.5Amp setup.
I had a 10A10 blocking diode on each string as I was direct connecting to a 144v nominal battery

10A10 diodes are normal silicon units, so 0.7v or 0.8v drop.
At 5A, they're getting way too warm (my temp gun was saying 180 deg C!!), and so a couple have now failed open circuit
Part of the problem was they were all in the same box, so didn't have much in the way of cooling.

I'm testing an MC4 plug in 20A 1000V diode, but at 5A, thats getting seriously hot too.
I thought that would fare better, as it can use the 4mm2 cable (that's only having 5A put through it) either side as a heat sink
It's still only a 20A10 solder into a MC4 connector housing.

The system voltage is too high for schottky diodes (or at least I can't find any that are rated to 200V).

If I could find a reasonably priced mppt that could take in 400v and put out 144v, I probably wouldn't be bothering with blocking diodes

[dRdoS7]

reviving an old thread here, but what blocking diodes do you guys use?


If I could find a reasonably priced mppt that could take in 400v and put out 144v, I probably wouldn't be bothering with blocking diodes

I have 4s4p, facing various directions: 3s tilted (E, N, & W), 1s flush (N). Some strings are 270W (IIRC), some 285W. Feeding a Midnite Classic 200.

I don't use any blocking diodes, just string fuses. Working OK.

dRdoS7

[T1 Terry]

reviving an old thread here, but what blocking diodes do you guys use?

on my setup I have 4s5p 175W 36Vmp, 5.5Amp setup.
I had a 10A10 blocking diode on each string as I was direct connecting to a 144v nominal battery

10A10 diodes are normal silicon units, so 0.7v or 0.8v drop.
At 5A, they're getting way too warm (my temp gun was saying 180 deg C!!), and so a couple have now failed open circuit
Part of the problem was they were all in the same box, so didn't have much in the way of cooling.

I'm testing an MC4 plug in 20A 1000V diode, but at 5A, thats getting seriously hot too.
I thought that would fare better, as it can use the 4mm2 cable (that's only having 5A put through it) either side as a heat sink
It's still only a 20A10 solder into a MC4 connector housing.

The system voltage is too high for schottky diodes (or at least I can't find any that are rated to 200V).

If I could find a reasonably priced mppt that could take in 400v and put out 144v, I probably wouldn't be bothering with blocking diodes

I use these attached to a tunnel heatsink that is also used for the solid state relays. A jaycar heatsink cut to the required length and slipped together and locked in place with a 4mm screw tapped into each slide about half way along, then a 50*C thermostat attached above the solid state relay/s either side and a maglev 80mm fan wired to the thermostats so the fan cuts in if the heatsink rises above 50*C.
Those diodes are a rectifier pair so they are wired opposite to each other to suit an AC wave pattern, a bit of a pain but sending each off to a solid state relay on either side seems to come out ok if you get a bit fancy with the wiring
We have to use the blocking diodes because the solid state relays have a flyback diode internally causing them to leak back to the solar array over night.

As far as the high voltage MPPT controller, the only ones I've found that weren't a crazy price were in the PIP 5048MG inverters. The Morningstar 600v version is up around $1,000, the whole PIP inverter/mains charger/80amp 450vdc MPPT controller all in one is less than that.

T1 Terry

Now I'll put the link in https://au.rs-online.com/web/p/rectifie ... s/0155513/

[T1 Terry]

I'll add a photo, that might make more sense

4 x SSR and 2 x diodes tunnel heatsink.jpg (285.5 KiB) Viewed 471 times

There are 2 fans in series on this set up because it is a 24v system.

T1 Terry

[antiscab]

Thanks Terry, that looks to be the way to go

Yesterday I got up on the roof and installed my plugin diodes (20A silicon diode internally)
I heat soak tested one at 4.5A continuous for 12hrs (at 0.65v drop, worked out 3 W), and it didn't seem heat up any more than it did in the first hour.
Though this was in a tin shed on a 34 deg C day, not under a hot solar panel on the highly reflective tin roof on a 48 deg C day.

If those fail, then I'll look at your beefy diode setup

I'm using the 144v battery that's in my car, so it's a shame I can't get the pip4048mg to work with it.
I'm not losing that much running without mppt, as the battery voltage sits just shy of the maximum power point anyway. I'll see how it goes in summer

[T1 Terry]

The solar array I had for my Prius was 290vdc open circuit, so no controller required because the 76 cell pack comes out at 3.8v per cell at the maximum open circuit voltage the solar array could provide, but of course there is no current at that voltage so it never got there. The 15 panels were at around their Vmp when each cell is at 3.35v, then the current started to taper off so the end of charge was virtually a trickle charge.
I got lazy and now just plug the 240vac 5 amp charger into our 7 mtr motorhome that has the 200Ah @ 24v lithium battery pack and 1.6kW of solar on the roof. The charging method is in no way efficient as it requires 1600w input to produce 1350w of charging into the battery. The only saving grace is the motorhome is just sitting there and the solar would be doing nothing if it wasn't recharging the Prius battery so we can drive home and back to work in ICE Kill mode

T1 Terry