New member from Durban

Member Introduction
21

This is a bad idea. Unless you have 3ph equipment, which you don’t, do not make a 3ph solar installation.

I’ll briefly go into some obvious why not to do its.

  1. There are 3 times as many things to go wrong. Why is this bad? It is bad because the entire 3 ph system stops when 1 inverter fails. If you had 3 inverters in parallel on a 1ph system and 1 failed, you could be back up and running in a hack on 2 inverters whilst the other is being repaired/replaced.
  2. Speaking of the replacement of a single inverter, these inverters can only be parallel or replaced with an identical model. And by identical, I mean identical chipset in the guts of the inverter. The same model inverter can have a range of chipsets. That chipset changes relatively frequently on certain models. In six years’ time, finding a replacement in a hurry is going to be a mission. All the while, those other two healthy inverters cannot provide a 3 ph supply. In a 1ph setup, there would be minimal downtime.
  3. 5kW /phase is quite limiting, exceed that on 1ph and the system is in overload. There is far more load combination flexibility if that was 15kW on a 1ph supply before you overload it.
  4. And now the converse of point 3. It is an inefficient waste of capital. These are very expensive bits of kit.
    Ideally, you want to buy an inverter that is capable of your requirements and not a watt more. Well, practically, that is difficult to gauge, but a 3 x 5kW set-up makes that nigh-on impossible. Something is going to be underutilized. Going back to my statement that the units have to be identical, that means you still have to put a 5kW inverter on a phase that may only use 2kW peak max.
    That 3kW worth of idle inverter cost is best deployed elsewhere.
  5. Cabling and switchgear requirements are more and, therefore, more expensive.
  6. It is understandable that, at this stage, you cannot visualize what a 15kW system looks like in terms of batteries and panels. Be it either 1ph or 3ph. There should be a number of panels and a number of batteries balanced with that 15kW of inverter capability. I think once you cut the coat to suit the cloth, your aspirations will tame.

There are probably, more good reasons, but I hope that is enough to convince you.

1 Like
22

Also, I would just like to add that in a 3-phase system (with a Multiplus or Quattro on each phase), the inverters aren’t really in parallel. Each one is master of its own phase, they merely sync their clocks so that the generated sine waves are 120 degrees apart. Setting up a 3-phase system (one inverter per phase) is significantly less complicated than setting up a parallel system with two or more inverters on the same phase. I mean, it’s not hard or anything, but from a technical viewpoint, there is very little to go wrong with a 3-phase setup.

Except, as Phil says, if you lose one phase, all three units switch off. Safety really, a lost phase is often a problem.

I mostly agree with him not to go 3-phase unless it is really needed. If your backup needs are modest, it is sufficient to simply put all the backups on a single single-phase circuit and install one Multi/Quattro :slight_smile:

1 Like
23

Thanks for the advice, the knowledge of the people here is why I am here. I can see that 3-phase is just a state of the incoming power, that I am not using (and won’t ever) use 3-phase appliances. I am pretty sure that it’s going cause issues if I just bridge the phases. Or am I - I know about a year ago fibre installers on the verge stuffed our municipal cable and we lost one phase. Their sparkie came in, it was in the evening, and just dropped a link from another phase in the actual DB and it worked without burning my house down. Later when the issue was fixed he just took out the link and we were back to normal?

I am perfectly happy to not have to buy 3 inverters, but at least two is my aim, that way I am not in the lurch should I lose one and it has to go for repair. How do I then split the combined output of two inverters (2 x 5kw, or 2 x 8kw) across the board which I will keep configured for 3 phases, in case I sell the house and have to revert back to the municipal power. Is it possible to have a single 16kw output from 2 x 8kw Multis?

24

You can have a split-phase setup (with an inverter on L1 and another on L2, but none on L3) and it will work perfectly fine. It is common in North America, in industrial setups they will have a 3-phase supply with 208V between phases (and 120V to neutral), but you only get two of the three legs to your property. It behaves the same as a split phase setup except the slightly lower voltage (208V vs 240V), so the American sparkies are not… uhm… phased :slight_smile:

But that kind of setup still suffers the other issue Phil raised, that it does limit you to a max per phase and that is slightly less flexible.

25

Welcome Trevor!

Funny we should me here!

Told you the real knowledge sits here. :grin:

1 Like
26

You have to prioritize your wants and needs.
There are three basic camps:

  1. People who want power during load-shedding.
  2. People who want to reduce their electricity bill.
  3. People who want to do both.

Firstly the cheapest solar is one that you don’t have to have.
People rarely have the budget to do 3 unless they play smart and reduce their needs and then break the issue down in terms of 1 & 2.

So:

  1. What loads do you NEED to stay on for the duration of load shedding?
    ( Not for two days, but for 4 hours. You must cover the 95% probability. You won’t be able to afford the 100%)
    Cost out if it is worth replacing higher power users with alternatives. (Led’s etc.)
    This is what you gear up for with a hybrid inverter and batteries using East and West PV flatter arrays.
    Overpanel about 180+% of hybrid inverter capability.
    You need enough PV so load-shedding does not draw on the batteries during the day.
    Probably looking at a 5kW Victron here. (Opinions on batteries are varied, but Pylontech seems to have been around for a while if you want to go off-the-shelf).

  2. What essential loads cost you the most money on your bill?
    Cost out if it is worth replacing power guzzlers with alternatives.( A++ appliances etc)
    This is what you gear up for using a PV inverter +CT, preferably using N PV arrays.
    Your power guzzlers should come on consecutively between 10 AM and 2 PM.
    Overpanel about 140+% of PV inverter capability. A PV inverter does not use batteries.
    Probably looking at 5-8kW here. ( Many options, Solis seems good value).

Do not think a 300W panel will churn out 300W all day. Don’t size PV for perfect conditions.
The inverters are secondary to the panels. You need to generate power first and foremost. Inverters only manipulate power.
Batteries are a hole in the pocket. You only want enough, not more than enough. Using battery power when ESKOM is available is (controversially) a false economy.

Edit: Buy panels by the pallet for better prices, and don’t buy odd or worse, prime numbers of panels. You won’t easily buy matching panels to expand in 6 months time if aesthetics are important to you.

3 Likes
27

Here comes @TTT (smoke wizard).

28

If we are voting … Go Single Phase in the whole house and automate the loads as needed.
Leave the 3P from the street and just bridge them.
Nice 8/10kw Multi and the rest of the list of bits :wink:

1 Like
29

But we have to be clear here. You cannot bridge them while still connected to the street :slight_smile: The bridging would be done strictly while islanded.

It is quite common where people use a single-phase generator to backup a 3-phase house. See @midnightcaller 's post here, for one wired exactly like that. You’ll note the bottom of the changeover has a blue and yellow wire connecting the phases together, but the top (the street side) is proper three phase.

30

I can easily be convinced to keep the 3-phase to the inverter(s) and then single phase to the DB. As said, no 3-phase kit, so it actually suits me to have all the inverter power from one source, rather than waste whatever headroom I might have on each of the 3 different phases.

The so question is - I would really like to not put all my eggs in one basket with inverters - can I get two 5kva, or two 8kva and have them use a single battery bank and supply a single feed to my DB, without the complications and added complexity of making sure I have identical firmware on each inverter.

My first idea:

12 to 16 x 555w Trina panels
1 x 250-100 MPPT
2 x 8kw Multis supplying a single output to the DB
2 x 5kwh Freedom Won batteries supporting the setup above.

31

On the battery side, everything stays the same whether you use 1, 2 or 6 inverters and 1,2 or 6 MPPT’s, you’ll only have a single battery bank.

You’ll just have to check the discharge specs of the Freedom Won packs to determine whether 2x of the 5 kWh batteries would be able to handle the current draw of the 2x 5 kVA or 2x 8 kVA Multies in parallel.

Just also check your solar panel calculations, 16x 550W will be 8800W.
The 250/100 MPPT can handle max 5600W of solar and though as discussed in other topics, oversizing a bit is good, I however think oversizing by over 50% will be a waist.
So rather look at adding another MPPT and splitting the array in half.

32

The Trina 555w have a VOC of 38.1volt and a SCC of 18.56amp

My thought was three parallel strings of 4 panels in series for 156v (4 x 39) and 76a (3 x 19). Or have I got my sums wrong here? And even if my sums are correct, the total watts over-rides the “apparent” capability of 250*100 (which is what you are saying and my recent reading confirms). Back to the drawing board on the number of panels and parallel/series setup,

Can I at least confirm that there is nothing preventing 2 x 8kw inverters providing a single output of 16kw (with the usual losses)?

33

Check it out here. See the oversize option at the bottom.

34

https://community.victronenergy.com/questions/135567/multiplus-ii-8kva-and-10kva-in-parallel.html

35

The MPPT 250/100 works a little different that you are thinking.
The 250 is for the max voltage before the smoke comes out. You can never go over this value.
The 100 is for the current (Amps), but not the PV amps. This is for the MPPT’s output (thus for the inverter and battery to use).
So the output current is at the battery voltage, not at the PV voltage. So if your battery runs at 51V that will give you 5100W (51x100) max, or at 54V it gives 5400W (54x100)
The spec sheet says 5800W for 48V batteries, so there are a bit of wigle room available (for a bit higher battery voltage for instance)
PS. this is why for 24V and 12V batteries to watts are less

2 Likes
36

I’ll try and give you an idea of how and why certain figures are relevant in the calculations.
Exceeding the voltage breaks down the insulation, which does irreparable damage.
So 250V is a hard limit. Open circuit voltage is what the MPPT will receive if it is switched on a cold day. So we will use the Voc we can expect on the coldest day in Durban in the calculations.

However, in terms of currents and wattages, we are talking about expected running conditions.
Those running conditions are stated in terms of “NOCT”, (normal operating cell temperature).
This doesn’t suit the marketing hype who would rather quote “STC” ( standard test conditions) values.
In reality, though your 555W panels, will only deliver that amount in a lab, they will deliver nowhere near that on a hot roof in Durban.

So to begin, the 250V capability of the MPPT should never be exceeded by the open circuit voltage of the PV string. But that isn’t the Voc as listed. That is the Voc on the coldest day in Durban.
So there is a Voc listed at a certain temperature, and a Voc temperature coefficient is also listed.

(I’ll use figures for a Trina 555W panel, I found on the web, you should check they are in line with your intended panels).
So, the Voc is listed as 38.1 V at 25deg C, ( with a 3% tolerance), which means it could potentially be 1.03 x 38.1V = 39.243V.
Then the temp. Coefficient of Voc is -0.25%/deg C and the coldest day ever recorded in Durban is 6 deg C, so we’ll use 5 degrees C as a safety margin.
This means that the panels could potentially operate at 25 - 5 = 20 degrees lower, and that would raise the coldest day Voc up by 20 x .25% = 5 %.

So 39.243V X 1.05 = 41.2052V.
As you can’t exceed 250V, the largest series string you can use is 250/41.2052 = 6.06 panels
So you could have 6 of those panels in a series string in Durban,( you wouldn’t get away with it in JHB as it gets colder there).

The next thing is the MC4 solar PV connections are only rated for 30A. If you tried to put 76A through them they would melt. Further to this 6mm copper cable is the largest cable that fits into these connections. Again this cable rating has to be respected.
For this calculation, you do not use Isc (@STC) because your MPPT will never see this current in reality.
It will see Impp(@NOCT) which is lower. This represents a truer reflection of what the panels will actually deliver. This is listed as 14.23 A. Bear in mind you probably will only get this occasionally as well.
So doubling up on that is still less than 30A.
So in terms of your array, you can have two parallel strings of 6 panels in series. This array will be the most efficient for a 250V MPPT.

To calculate an optimistic expectation of the power you can expect from such an array use Pmax Wp (@NOCT), which is 420W /panel.
12 X 420 W = 5040 W.
This means a 250/100 Mpp will be capable at a battery voltage of 51V.
(51V is for a 15-cell battery, you could even get more leeway using a 16-cell or higher battery).
Empirically, I have satisfied myself that if you are not using ideal North facing arrays, you would probably get away with a 250/85, or even maybe a 250/70.

I want to go on to say, that this size array is still hopelessly under-sized for a 2 x 5kW inverter capability. You want (at least) three of this sized array to achieve a balanced system.

1 Like
37

Thank you, this is an excellent analysis that even the layman like me can understand. I’ll be checking to see exactly how many panels I can get up on my only viable roof, north facing. 14 looks easily doable, I’ll use your calcs to make sure I’m not breaking any rules.

1 Like
38

E + W roofs are also viable (if you have them), and for the security of supply applications, a combination of both can be better suited than just N-facing alone.

1 Like
39

A solar installers nightmare. Double story home, very small main residence roof with East/West. The North facing roof is a separate garage and garden flat 50m away from the main residence. I’d have to pull 50m of solar cabling in one or another direction. The garage/flat is the location best suited for the kit, I have no real space in the main house. I’m going to pull AC cabling from the meter to garage and garage back to the DB. It’s a mission but in my mind the only way to do it.

40

A quick diagram of the layout would help. I’d need to know which phases go where.
From what you’ve described, I think it can be done with existing cabling.