Bobster.

Errrm.... I think this is above my pay grade 😂

Also we're talking about settings that my inverter doesn't have, so I can't even read the fine manual and comment.

What I do have though is parameters for the grid. This specifies a range of voltages (and frequencies) for which the inverter deems grid to be available and of sufficient quality. Somewhere south of 190V my inverter will disconnect. It still monitors the grid but won't interact with it. If the grid voltage comes up enough and stays there for long enough then the inverter reconnects.

But this may not be very helpful, we are talking about different brands of hardware with different programming and design.

This sounds like a risky tactic to me. I've heard about meters doing this, but how is a computerised billing system going to deal with this?
Last month closing = 123456
This month closing = 123111
Oh wow! The meter must have gone all the way around, zeroed and then gone forward again. The system will not stop to consider how you might have used so much juice in the space of a calendar month.

If the feedback amount is small then you might get away with it (EG would be 123489 instead of 123567), but eventually you will end up in the situation I describe. 

If these are the 145V max PV models, then the latest firmware seems to be 72.05, available here.
But Kodak seem sloppy with model names, not distinguishing properly between the King and King II models. The latter have 500V max PV voltage. Check the label on the side of the inverter.
Sorry for the late reply.

Hmmm.... one can currently get new 10/8s for not much north of 40K.

EG https://powerforum-store.co.za/collections/batteries/products/freedom-won-lite-home-10-8-lifepo4-battery-n

The inverter can't handle this really. Charge limit for the hardware is 100A, and the Goodwe agents won't set it past 80A as per - they say - an instruction from FreedomWon. Though if the hardware can't go past 100A, there's not a lot more that can be done. The bus bar is the main piece of the puzzle though. Batteries connect to this, and the bus bar feeds the disconnect. The two batteries would have to talk to each other as well.

FW want a minimum distance between the batteries, so I need to measure to see if I can meet that requirement.

You may have to re certify. If your system is registered then you would need to de-register it and register at the new address - a system is registered as an inverter (identified by serial number) at a physical location, and the registration covers all the electrical connections.

You'd probably have to "un split" the DB (cost) and may have to patch walls where the equipment is removed. Those costs are unlikely to be close to the cost of a PV system, but you should still figure them into your costings.

Can you check if inverter 3 is missing a capacitor across the PV inputs? MKS IVs are known to be sensitive to this, usually affecting the display, but it might conceivably trigger other issues.
There is a small file in the downloads section with slightly more detail. Be aware of the safety issues if you take this on yourself. Big high voltage capacitors can cause nasty shocks and/or damage underwear.
Edit: Since fault code 72 is to do with current sharing, it may be to do with the length and/or resistance of the AC-out cables. If these are significantly different (you mention a 1 metre difference), it could be worth lengthening the cables from inverter 1 to the joining point, so that the lengths are more even.
It's also possible that the current sense circuit for inverter 3 (or others) has a bad capacitor that would normally smooth out noise, for example from the PV input cables.

Lol. Normally the people who refer to me as a "gentleman" are the police.

One method that has been suggested to me is to look at what's being sold second hand. The theory behind this being that if people are getting rid of them it's because they are unhappy with them. You could refine this by finding what brand of inverter the seller has. I used to have Revovs, they didn't get on well with my inverter, so it's a good bet that Goodwe owners are more likely to dump Revovs than Victron owners are.

This isn't scientific, but it may reflect real world experience.

And the PV system powers the pump, so you have a constant supply.
I have a 2.5Kl tank. I calculate that will get us through a 10 day outage if we're careful. The longest we have had so far is 4 days.

I can give some general answers. I'm not an Eskom client, but I have had my system registered.
You live in Cape Town. There must be companies there doing registration & especially checking for compliance. Because COCT requires registration ahead of installation. You should be able to find somebody. Up here in Gauteng there are companies that deal in registration, I'd expect Cape Town to be better off in that regard because of the greater demand (which comes because of enforcement of the regulations).
Once you find that person I would think that if they're doing a decent job they would check your system for compliance and advise you.
One of the reasons Eskom will want to replace your meter is so that you can sell back to the grid. To do that you need a meter that can differentiate what you sell from what you buy and thus allow accurate billing.

The arithmetic is straight forward: If the chance of a stand alone inverter failing is 1 in 100, then with two in parallel
1) The chance of either inverter failing is now 2 in 100.
2) The chance of losing both inverters at the same time is now 1 in 10000

If it's a logical OR (Inverter A fails or inverter B fails) then we deal with the sum of the probabilities. If it's a logical AND (A fails AND B Fails) then we deal with the product of the probabilities.
If A fails, then the chance of B failing is now 1 in 100 as we are down to one inverter.
Having them in parallel increases the chance of one of them failing, but reduces the chance of a total loss of backup.

Interesting to note too that although the BMS can instruct the inverter as to what it wants, and a user like me cannot change that, Goodwe can impose an absolute charging limit that overrides anything the BMS requests.

Ok, lots of off-topic postings, let me muddy the waters waters by throwing in my 5 zim-cents worth...
If all you need is more kW power from your inverter, but not more solar production options, then two 5kWatters are fine, but... looking into the future and maybe even now, where you may want to add more solar panels to produce more power, the 5kWatters are a bad idea, they are limiting the current on the PV inputs such, that quite a few panels that are available these days won't foot the bill and you may struggle to find panels that will work well with the Sunsynk/Deye 5kW inverters. The 8kWatters are a better option and looking at Solis should not be off the table either.

I agree that having a spare in stock rather than in use is a good practice (if we can afford it).

Otherwise I fall between two stools.

If I have two inverters, same make and model, same age, in parallel then I increase the number of components and so the chance of an inverter failing actually increases for failures due to component failure.

But the likelihood of losing all functionality will decrease. The two are unlikely to have the same component failure at the same time. One inverter will fail, but the other remains in service and so I have some backup.
Caveats:
1) All odds are off for failures caused by something external EG power surge. Though Bob poking around the insides of the inverter is also external, and hopefully after I blow up one I learn my lesson and let the second be.
2) The two inverters would not be backing up the same circuits. So I might keep the pool pump and outbuildings going, but lose supply to the main house. But somebody who knows what they're doing can swap the connections around, or some sort of change over switch could be put in place. I would lose backup on some circuits - unless we can add the outputs of the two inverters so that a failure of one inverter will give us same circuits backed up but we have to be more careful with what we switch on.

Your proposed approach aligns closely with recognised best practices for lightning and surge protection in photovoltaic systems in South Africa, particularly in regions prone to frequent thunderstorms.
We as equipment suppliers strongly recommend the installation of Type 1 or combined Type 1+2 surge protection devices (SPDs) such as those from reputable brands like DEHN HAGAR CHINT on both the DC and AC sides of the system.
These devices are specifically rated to handle high energy impulses from direct or nearby lightning strikes, offering superior protection compared to Type 2 SPDs alone, which are more suited to induced surges.Earthing remains fundamental.
Install a dedicated earth electrode 1.5 m copper clad rod or multiple bonded rods at the PV array, connected via heavy-gauge cable typically 16–35 mm² copper to the panel frames, racking, and SPDs.
Aim for a ground resistance of less than 10 ohms, as commonly required for compliance and effective surge diversion. A common equipotential bonding conductor 6–16 mm² should then interconnect this to the inverter, main distribution board, and building earth system.
This creates a single, low-impedance path to ground, minimising dangerous potential differences during a strike and ensuring surge currents follow the path of least resistance away from sensitive equipment.
For enhanced protection, incorporate multiple SPD installation points one set at the array or combiner box particularly with longer cable runs and another immediately before the inverter on the DC side, complemented by AC side protection.
This staged or cascaded arrangement progressively diverts energy, significantly improving system resilience in high lightning areas.
In locations with elevated risk such as exposed rural sites or high keraunic level zones consider external lightning protection measures, including air terminals (lightning rods) with down conductors, in accordance with a risk assessment per SANS 10313 or IEC 62305.
This is typically reserved for more extreme conditions and may not be necessary for standard residential installations.
Regarding insurance and compliance
A valid Certificate of Compliance (CoC) issued by a registered electrician is essential for grid-tied or hybrid systems, and insurers frequently require it for lightning related claims.
The CoC should document key elements, including measured ground resistance (ideally smaller than 10 ohms), SPD installations, equipotential bonding, and adherence to SANS 10142-1 (and related standards such as NRS 097 for grid interconnection).
Without properly certified documentation, claims are often declined.
In summary, this layered strategy combining high capacity SPDs, robust equipotential earthing, multi stage diversion, and full certification provides the most reliable defence against lightning induced damage.
It far outperforms reliance on manual disconnection of the DC breaker alone and positions the installation well for long term reliability, safety, and insurance acceptance.
I am sure there are a lot of installers and engineers on the forum that can add their opinions however some of these topics are hotly contested and debated however we prefer to follow best practice advice bound to the SANS Regulations to our clients.

Is it not the case that "cycle" is defined by, in this case, FreedomWon and not by the BMS? They will guarantee you so many cycles and then the guarantee will define what a "cycle" is.

It's fine. Thanks. I did get a couple of shocks on my work bench days. We all did. I did know to use a screwdriver to short the HT lead on a CRT to earth before poking around there. We also used to amuse ourselves by inflating a paper bag and bursting it behind somebody's head just as he stuck his fingers into whatever he was fixing. This passed for humour.

I know there is usually a good reason for things like having an isolator as well as a breaker, even if it's not obvious to me. So this was a chance for me to learn something.

Thanks again for the good answer.

I have a heat pump and PV. In fact I had a heat pump before I had PV.

I got the heat pump because the geyser was not near the north facing roof, so the installer recommended a heat pump.

This has the advantage that you note: I still get hot water on not-so-sunny days. It also has lower power demands than a regular geyser element, so I can run it on the backed up side of my solar system. This means that during load shedding I always had hot water.
My heat pump is not integrated with the geyser. There may be advantages to this. If (when) my geyser goes pop then it's a straight forward replacement.

The pump has been in service now about 13.5 years.

The situation: We are two adults, nobody else in the house. We have set routines. I run the pump twice a day: 6:00 for one hour and then 12:00 for three hours (that's the window I give it to run in, it doesn't run for 3 hours).

I have a blanket (Isotherm) on the geyser and cladding on the hot water pipes. After the midday run my wife is able to take a hot shower at 9, even 10 in the evening.

This is now a good combination for me. I am heating water free, gratis and for nothing. Every day. I prefer it to solar water heating. At our previous address I had a solar geyser. One problem was that it would overheat in the summer and a pop-off valve would let lots of water out. That didn't make me happy. Also water temperature varied across seasons - pleasant but not hot in the winter, too hot in the summer. Sure I can mix some cold water in, but I could see a situation where a person, especially a kid, could get caught by surprise. Now I have a pretty constant temperature throughout the year, the pump won't heat past 60 degrees. With the solar, to get the most bang for no bucks, you have to fit in with its schedule - so showers in the afternoon and early evening.

I run my pump twice during the day. I don't run it at night. If I did then it would draw from the battery, but then so would a geyser element (and more current and for longer).

We also have got into the habit of short showers once a day. That gives whatever water heating system we have less work to do.

https://www.dailymaverick.co.za/article/2026-02-24-sas-power-market-flips-as-eskom-monopoly-ends-municipalities-face-squeeze/

Agreed. I had a heat pump before I had solar, so my strategy was different. I figured out that I could have hot water despite load shedding. But load shedding isn't with us any more, and so that's no longer a strong card to play.

I'm another long time heat pump owner (so my pump is not as efficient as those you buy now) and I disagree with this. It is true that the pump is slower in the winter BUT I can run on it the backed-up side of my PV system. Because of that I think it's a good option. I am not using grid electricity to run my pump.

Also I checked the data from my PV system for July. In the morning the pump runs for an hour - that's what I allocated it on it's own timer. I don't check the temperature every time, but I don't recall getting into the shower and thinking that it's a bit cold. The second run starts at 12:00 and the timer will let it run until 15:00, but actually and in the winter, it runs for about 45 minutes

Less than a year ago I paid 30 grand for
5 new panels - placed on existing mountings
4 of the old panels to be moved to another string
Proper earthing of the two strings
Two PV fuses so +ve and -ve are fused
COC
Replacement of a faulty surge arrestor
That was with a R1K trade in for two of the old panels.
The five new panels and all other new parts were supplied by the man who did the job.
It pays to shop around. But better still is to get recommendations because, as @Denns points out, the cheapest job isn't always the best job.

Thanks for the explanation.

Surge protection is about voltage differences between earth points.
If: Main DB earth and pool earth rise to different potentials during a lightning event that voltage difference is what damages equipment. If the loop resistance between main db and sub db is negligable then jou are good. From traveling back to the sub through the cable is possible but unlikely.
That’s why good bonding and low earth resistance matter more than just adding devices.