TaliaB

COCT only cares about changes that affect the grid-facing characteristics of the system (AC side). Pure DC-side changes do not affect the grid and therefore usually don’t trigger re-registration or a new COC.
Scenario 1: Add more panels to the existing inverter (within inverter limits)
No re-registration required
No amended COC required
Why:
Panels are on the DC side. Inverter model, rating, and export capability remain unchanged
Anti-islanding and grid protection are unchanged
Important conditions:
Must remain within inverter:
Max PV power
Max Voc (cold conditions!)
MPPT current limits
No AC wiring or protection changes
This is why COCT does not require an amended COC for DC-side-only upgrades the COC covers the electrical installation, not panel count.
Scenario 2: Add a second inverter with additional strings:
Re-registration required
New or supplementary COC required
Why:
Inverter capacity changes
AC wiring and protection change
Grid connection characteristics change
You’ll need: Updated SSEG application
Updated Single Line Diagram (SLD)
Export limit confirmation
Cable upgrade if existing wiring (e.g. 6 mm²) is no longer adequate
This is a material system change and must be declared.
Bottom line:
If the AC side and inverter configuration stay the same, COCT doesn’t care.
Once you change inverter count, AC wiring, or export capability, paperwork is required.
This is why DC-side upgrades do not need an amended COC, even though many people assume they do.

Should you decide to parallel the 2 above-mentioned packs Ensure before you do the parallel connection that both batteries was charged individually to 100%Soc. Also check the terminal voltage before connecting in parallel and should not have more the 0.5v diffrence. Then pre- charge the inverter and connect the paralleled bank and charge both together to 100%Soc.

Asked and answered by @hoohloc

Being part of a BESS installation in La Réunion island no stone is unturned regarding safety. A company called Akuo Energy was tasked with the building of Lfp battery rooms and enclosures..
Primary focus was on the prevention of a thermal runaway catastrophic event within Lfp battery enclosures. When thermal runaway starts propagating the first line of defence is the detection of Hydrogen gas. The failure progression happens in distinct stages, moving from silent chemical breakdown to catastrophic fire or explosion.
The Pre-Runaway Phase (Off-gassing): As the cell heats up and reaches roughly 150 to 180 deg due to an internal short or abuse the solid electrolyte interphase (SEI) begins to degrade. This breakdown releases Volatile Organic Compounds (VOCs) and hydrogen gas before the full runaway sequence takes over.
To mitigate full blown fire or explosion Hydrogen sensors are used that triggers flame explosion proof extractor fans and AOV's( automatic opening vents)to ensure the Hydrogen and air mixture does not reach critical levels.
For us with small scale lfp installations your best friend would be well ventilated battery areas with moderate charge and discharge curves. Hydrogen sensors could be fitted as a early warning but that would be an expensive exercise. I will do the research and see what i can find in South Africa in the form of pure diatomic hydrogen gas warning systems.

Being part of a BESS installation in La Réunion island no stone is unturned regarding safety. A company called Akuo Energy was tasked with the building of Lfp battery rooms and enclosures..
Primary focus was on the prevention of a thermal runaway catastrophic event within Lfp battery enclosures. When thermal runaway starts propagating the first line of defence is the detection of Hydrogen gas. The failure progression happens in distinct stages, moving from silent chemical breakdown to catastrophic fire or explosion.
The Pre-Runaway Phase (Off-gassing): As the cell heats up and reaches roughly 150 to 180 deg due to an internal short or abuse the solid electrolyte interphase (SEI) begins to degrade. This breakdown releases Volatile Organic Compounds (VOCs) and hydrogen gas before the full runaway sequence takes over.
To mitigate full blown fire or explosion Hydrogen sensors are used that triggers flame explosion proof extractor fans and AOV's( automatic opening vents)to ensure the Hydrogen and air mixture does not reach critical levels.
For us with small scale lfp installations your best friend would be well ventilated battery areas with moderate charge and discharge curves. Hydrogen sensors could be fitted as a early warning but that would be an expensive exercise. I will do the research and see what i can find in South Africa in the form of pure diatomic hydrogen gas warning systems.

Wise words indeed - but there are unfortunately many in this industry that either don't know, or don't care because they don't know, of the risks associated with our solar power systems.
In my opinion, and based on my own experience gained as a "home user", the industry should drive for improvement in 4 main areas:
Education: I thought that I was "reasonably" educated in the theory and application of PV systems (at least on a "home owner" level), but never knew about the substantial risk to life resulting from a thermal overload and rupture of a battery as a result of the highly toxic vapour released during such an event. At no time during my discussion with various sellers and installers of systems (whilst I was "shopping" for my system nearly 2 years ago) did ANYONE mention this chemical risk - and I'm sure that I am not alone in this... In my opinion, the manufacturers, suppliers and installers of lithium-ion batteries have a moral obligation to inform their (potential) clients of this danger - even though there is a very low risk of such an incident happening.
Installation: Again, speaking from my own experience - my system was installed compliant with all the requirements of the SANS electrical code... Four massive isolators/fuses were installed between the two batteries and the busbar boxes to interrupt current either deliberately by being opened, or by blowout of the 250A fuses on both positive and negative wires leading to the busbars. The high-voltage DC from the PV panels leading to the inverter is routed through a dedicated DC distribution board (enclosed in it's own box) with surge protectors, fuses, and DC disconnection switches - all compliant with code... The problem (that became clear now that I've done an informed risk assessment), is that both the DC isolators from the batteries, the AC and DC DB boxes, and the trunking with all the fat DC cables in them, are all: a) made of plastic; and b) located directly above the batteries. Now, I've come to know the installer quite well during the installation, and neither he, nor the master electrician that issued the COC, were negligent in anything electrical - but these DB boxes and isolators were installed in the worst possible place to actually reach them in case of emergency if a battery did rupture and started off-gassing, or was actually burning. This, I'm guessing, was as a result of not knowing and/or fully appreciating the inherent chemical and fire risks associated with these batteries! So, the installer (in my case) was not properly trained in these safety aspects by Industry (e.g. no warnings in the installation pamphlet supplied by the manufacturer of the batteries), and a "non-electrical" risk assessment was also not part of the COC inspector's duties...
Design: Knowing what I now know, I would have included more passive fire protection measures, even though it would have increased the installation costs. I will certainly look into installing a "fire-proof" enclosure around my batteries with a "venting" chimney leading up from this enclosure, through the ceiling of my garage (where the system is installed), and through the gable wall, so that any out-gassing fumes can be safely vented away from the rest of my house - this should not be too expensive, and should at least protect the house, and it's inhabitants, from this very-small-chance-of-happening-but-with-extremely-serious-consequences event. Obviously, I would also not have allowed the disconnects and DB boxes to be installed close to, and above, the batteries - just to save on some cabling cost...
Operation: Don't abuse the batteries! I've seen frequent statements on this Forum where people are "fast charging" their batteries at their maximum rated current, or regularly "deep discharging" the batteries until they cut off during loadshedding events, without any concern regarding the thermal and chemical stresses that such actions apply to the batteries. This is probably due to the industry marketing tactic to rate their batteries as "100% dischargeable", "fast charge enabled", etc. without adding at least a warning of potential consequences.
Well, that was a long one - but I think this matter warrants some detailed discussions by both "owner/users" and installers.

Incident closer to home. In all of the information gathered on this post what is or was the root cause of the thermal runaway events? Take for instance the sealed container, what was the temprature of the battery packs inside the non ventilated container or was it ventilated?? In all of the videos the exact cause of LFP failure is not disclosed or diffucult to pinpoint. What is the failure rate of LFP 1 out of 40 million ?
How to prevent thermal runaway in our domestic setups. Basically do not stress abuse or damage the cells apart from that nothing in life is certain misfortune lurkes in every corner of our lives in diffrent shapes and forms. All that we can do is to endeavor to be safe as humanly possible.
FIRE OPS SA
FIRE OPS SA - Batteries catch fire at Standard Bank offic...
“Fire Ops teams were dispatched at approximately 09:00, where they encountered an intense blaze inside a container housing two 300kWh lithium-ion battery racks,” said Koekemoer. “Despite nearly two ho

Some important information about LFP thermal runaway from Battery University.
https://www.batteryuniversity.com/article/bu-304a-safety-concerns-with-li-ion/

You might want to have a look at the Elon Smart system from Kwikot . Below link for the system setup. I have installed a few and the work great. 3 x 550w panels in series will work quite well.

Solis Hybrid Inverter 5KW S6-EH1P5K-L-PRO does support generator integration and effectively has a configurable GEN/Smart Port capability automatic generator start/stop, generator mode peak shaving, smart port functionality.
So in your earlier kettle example: 2 kW inverter generator connected to the Solis: kettle drawing 2.8–3 kW if you have PV available and/or battery …the Solis can supplement the shortfall from: battery, PV, or both simultaneously.
It behaves much closer to a Sunsynk/Deye hybrid than to a traditional off-grid inverter.
The generator does not need to carry the full kettle load alone.
The inverter can “assist” using DC-side energy.
PV production during daytime reduces generator loading automatically.
The only caveat same as with Sunsynk is that tiny inverter generators can sometimes have voltage sag, frequency drift or react slowly to sudden step loads if that happens, the Solis may disconnect from the generator input temporarily. Victron systems are still generally considered the benchmark for weak-generator handling through "Power Assist and ESS, but the newer Solis S6 hybrids are much more capable than older hybrids in this regard.
SANS compliance for the generator installation itself still needs to be considered separately. Generator output -dedicated breaker/isolator - inverter GEN input. Proper earthing conductor from generator frame to the installation earth system. Neutral-earth bonding handled in only ONE active location at a time. As soon as utility power( Eskom is interrupted) the inverter will create the earth/neutral bond on output so no need to bond the generator as GEN input is not seen as utility input to the inverter input.

Solis Hybrid Inverter 5KW S6-EH1P5K-L-PRO does support generator integration and effectively has a configurable GEN/Smart Port capability automatic generator start/stop, generator mode peak shaving, smart port functionality.
So in your earlier kettle example: 2 kW inverter generator connected to the Solis: kettle drawing 2.8–3 kW if you have PV available and/or battery …the Solis can supplement the shortfall from: battery, PV, or both simultaneously.
It behaves much closer to a Sunsynk/Deye hybrid than to a traditional off-grid inverter.
The generator does not need to carry the full kettle load alone.
The inverter can “assist” using DC-side energy.
PV production during daytime reduces generator loading automatically.
The only caveat same as with Sunsynk is that tiny inverter generators can sometimes have voltage sag, frequency drift or react slowly to sudden step loads if that happens, the Solis may disconnect from the generator input temporarily. Victron systems are still generally considered the benchmark for weak-generator handling through "Power Assist and ESS, but the newer Solis S6 hybrids are much more capable than older hybrids in this regard.
SANS compliance for the generator installation itself still needs to be considered separately. Generator output -dedicated breaker/isolator - inverter GEN input. Proper earthing conductor from generator frame to the installation earth system. Neutral-earth bonding handled in only ONE active location at a time. As soon as utility power( Eskom is interrupted) the inverter will create the earth/neutral bond on output so no need to bond the generator as GEN input is not seen as utility input to the inverter input.

On VRM portal create a Widget to show you battery SOC vs Battery current vs Battery voltage. Also ensure your bms is communicating with the system. Then on VRM check under Battery details for the highest and lowest cell voltage.

Yeah, I saw issues just like this in the past - RCD on the input side of inverter tripping when switching between modes. Not on every try, but pretty often.
Just to emphasize, these are RCDs, not MCBs. So it's not a spike or overcurrent, but in most cases the RDCs trip because of imbalance in current passing thru L vs N line. Which is normally caused by leakage current, or by shorting PE+N after the RCD.
Root cause:
Internal grounding relay (PE+N) of the inverter must be closed in the off-grid mode, but open in the grid mode.
When inverter is switching between modes and connects AC input to the grid before disconnecting internal PE+N relay, then RCD trips.
Since it is unsafe to run off-grid mode with the internal relay open, the inverters are trying to minimize the time with PE+N relay open - and sometimes it happens that AC input is already connected while the PE+N relay is not fully opened yet. Then RCD trips.
For inverters running in parallel it's even worse, as their mode switching logic is much more complicated.
The behavior depends on the logic incorporated in the inverter's firmware. Some models are doing OK, others are causing troubles.
The working solution is to put RCD at the inverter output, not at the input. RCD installed at the inverter input, especially when more inverters are running in parallel, is a free ticket for these troubles.
PS: For example, in my country, RCDs are mandatory for output circuits, like light, sockets, domestic appliances, water pumps, in order to protect users. If these circuits/devices are being fed by the inverters, then there's even more strict requirement - must use Type B RCD, the one that can detect DC currents. But, AFAIK, we have no requirement here to use RCDs for inverter inputs - just MCBs are sufficient.

Switch OFF lithium/BMS communication and use “Lead Acid/User Defined”.
Manually enter safe lithium voltages ONLY as a short test. If the battery immediately starts charging the problem is definitely BMS communications or permissions.
This is one of the fastest ways to isolate the fault.
Based on the symptoms, my suspicion order would now be battery BMS charge inhibit
1.CAN/RS485 comm issue after firmware update
2.Max charge current accidentally set to 0A
3.Corrupted TOU/force-charge state
Less likely: actual PV issue
The fact that PV instantly ramps when the pool pump starts is actually very good news the solar side itself appears healthy.
Then Fully restart BOTH battery and inverter
Not just the inverter.
Proper sequence:
Turn off PV isolator
Turn off AC breaker/input
Turn off inverter
Turn off battery breaker/power
Wait 5 minutes
Then restart:
Battery first
Wait until battery fully boots
Inverter second
AC input
PV last
This often restores CAN/BMS comms.

Yes, it should technically work if the Solis firmware supports frequency-watt derating (frequency shifting / droop control) in island mode, and many people have AC-coupled Solis units to Victron systems successfully. But I would definitely verify the exact Solis model + firmware because some regional Solis firmware versions historically did not respond correctly to Victron frequency shifting. Also ensure the PV inverter sizing (factor 1 rule) is sensible relative to the 3 × 5kVA Multis and battery bank.

Below Relay configuration for Neutral/Earth bonding using utility(Eskom) to drive the relay coil. Use a 1 amp fuse on the live leg of the relay coil.

Not familiar with Macsell.store never bought from them. The Dyness lfp is good quality.
Bought from Solar Warehouse professional service, staff has good technical training on items they sell and will go the extra mile to sort issues out. I can recommend them.They also sell the Dyness DL2.5 25.6v 100ah Lfp battery. See link below.
Solar Warehouse SA
Dyness 24V 100Ah 2.56Kw DL2.5 Lithium Battery
Dyness 24V 100Ah 2.56Kw DL2.5 Lithium Battery Dyness DL2.5 is a good alternative for lead-acid batteries and a perfect match for on/off-grid applications in areas with limited or no grid access. It is

@Aartappel your problem is quite a generic one and I have seen a few different solutions that the farming community is deploying...Direct Solar panels on VFDs, some farmers creating micro Grids large in size and then installing their own MV transmission lines and then disconnecting the various Eskom incomer points. I have at least 3 users completed these projects & have done it in a staged approach but the savings and the benefits are there to witness. Those unreliable expensive Eskom points are worth shutting down & pocketing the ongoing costs into your own infrastructure.
I have been involved with especially ATESS and more recently over the last 2 yrs Megarevo Transformer based Hybrids and AC coupled Grid ties and DC coupled ATESS topology. The largest plant at 1.5MW carrying a base load of 700kW is in a factory with massive HVAC and high inductive loads. The Atess equipment is really robust and honestly can take a heavy punch with large motors starting and stopping...
Its just my personal opinion that the High frequency non transformer based inverters won't last long for the irrigation environment or SAs farming environment because we have harsh conditions and we require equipment that will last the next decade that is serviceable in the field.
If you need basic advice and user cases please PM me. I can definitely point you in the right direction as well as share plants with you that have proven themselves in the field. The farming community are ripe to reap the rewards of solar power & with Diesel pricing very volatile id imagine it's going to accelerate...

Hence the reason i posted this thread on specifically that item. Imported JA 550w per container(620 panels)~ R1.50 to R 1.70 per watt depending on International bulk( FOB) prices. So yes you can't even import bulk Tier 1 panels for that price.🤔

No you cannot just install a 300–350 kW hybrid inverter system behind a 200 kVA Eskom supply without formal approval and likely network upgrades.
You mentioned: Supply: 200 kVA
Proposed inverter: 300–350 kW (~300–350 kVA) Immediate problem: You are proposing generation larger than your point of supply that triggers Nersa registration required and Eskom connection study.
Not a simple SSEG anymore moves into embedded generation project territory.
In real projects like yours (irrigation / farms)
Option A What most guys do (legal route) is upgrade supply to 500 kVA or MV then install 300–500 kW solar.
Option B Hybrid workaround (semi-legal if approved) Keep 200 kVA supply and install large PV. Use zero export + strict control, but still requires Eskom approval.
The second option if you don’t want to upgrade to 500 kVA Eskom supply you can go off grid(remove point of supply) and install what you want(need) with backed up supply in mind.

Your comments mixes a few practical truths with some assumptions that don’t always hold up technically or economically.
I agree with the first part smallest battery that meets your actual use case usually gives the best ROI. Especially in grid-connected systems, batteries are often the most expensive kWh you’ll ever “buy,” so oversizing them for things like geysers or daytime loads doesn’t make much sense when you can rather push that load into solar hours.
Where I differ is the jump from 10kWh to 20kWh being a “problem.” That’s more a system design and brand limitation than a universal truth. Many modern rack systems (Pylontech, Hubble, Dyness, Freedom Won, etc.) are inherently modular at ~5kWh per unit, so expansion is incremental by design. If someone installs a monolithic 10kWh battery, then yes they’ve locked themselves into larger expansion steps, but that’s a product choice, not a battery principle.
On the second-hand argument:
In theory, 5kWh units might be more liquid but in practice, the second-hand battery market is still very immature, and buyers are far more concerned about:
●Cycle count
●State of Health (SOH)
●Brand reputation and BMS compatibility
A 10kWh unit with good telemetry and warranty traceability will often be easier to move than multiple smaller units of uncertain history.
Also worth noting: more small batteries = more parallel strings, more comms complexity, more potential imbalance issues over time. There’s a real engineering trade-off there, especially in systems that already push multiple parallel packs.
The “phone two friends” comment is funny, but practically speaking, installation logistics shouldn’t drive system design. Most 10kWh-class batteries are still manageable with proper handling, and installers deal with this daily.
Get yourself one of these:

For me the bigger picture is this:
Size the battery for night load and outage requirements only use solar + load shifting for high-energy loads (geysers, pool pumps, etc.)
Keep the system modular where possible, but not at the expense of unnecessary complexity. If anything, I’d argue the real optimisation is:.Spend on panels first, battery second especially if the grid is available.
That’s where the real ROI difference sits.

To answer @MuneebK question 1 x 10.6kWh or 2 x 5.32kwh batteries? Installing 1 battery module makes more sense less wiring less components much simpler.
As for the rest of the discussion my 2 cents. Try not to use high resistive loads(geysers) from your batteries.
Let us compare two very different cost models: PV generation vs stored energy.
Additional solar capacity (R/Wp) is still far cheaper per kWh produced over its lifetime than LiFePO⁴ storage (R/kWh usable).
Using batteries to run a 2–3kW resistive load like a geyser is one of the least efficient uses of stored energy (high discharge rate + unnecessary cycling).
If utility/grid is available, it generally makes no financial sense to size battery capacity for occasional cloudy periods just to support non-essential loads like water heating. That’s an expensive way to avoid a relatively cheap backup source(Eskom) used a few days a year. A better approach is load shifting + thermal storage: Heat the geyser during peak PV hours (10:00–15:00) and set thermostat to ~70 °C to increase stored energy capacity. Use a geyser blanket + pipe insulation to reduce standing losses. Let stored thermal energy carry evening demand (mixed down at the tap).
This effectively turns the geyser into a low-cost energy storage system, avoiding battery cycling and preserving stored electrical energy for essential loads..Also consider load staggering only one geyser active at a time..Prioritise based on SOC or excess PV.
Hierarchy should be:
1.Use PV directly
2.Shift loads to sun hours
3.Store as heat (geyser)
4.Use batteries for essentials
5.Use grid as fallback
Oversizing batteries to run high-power thermal loads at night is usually a capital inefficiency, not a system optimisation.

You ain't seen nothing yet 😂
You’re 100% right on one key point not everyone is solving the same problem.
If your primary goal is resilience and independence from Eskom especially with cable theft, outages, and general instability then oversizing inverter and battery capacity is a completely rational decision. In that context, ROI isn’t just rands per kWh, it’s availability, reliability, and control, which are hard to price but very real.
Where I’d challenge the approach slightly is on how that oversized system is then operated. Using batteries to deliberately dump energy into a geyser at 04:00 to “make space” sounds logical at first, but technically it’s still one of the least efficient energy paths. You’ve already paid a premium to store that energy in LiFePO⁴ you then cycle it (adding wear) only to convert it to heat which could have been done directly from PV the next day
Even with “excess PV” in mind, most systems can be tuned to absorb surplus during the day instead: Raise geyser setpoints during solar hours (e.g. 70°C with good insulation) and use staged or priority-based load control (one geyser at a time)
Shift workshop loads where possible into sunlight hours on weekends or flexible days. That way you avoid unnecessary battery cycling, preserve battery lifespan (which is a major capital component) and still maximise PV utilisation.
On the “creating storage space” argument in a properly sized system, that situation should actually be rare. If you’re regularly hitting full batteries and clipping PV, that’s usually a sign the system could benefit more from additional daytime loads or even slightly less battery relative to PV
Your use case (lathe, welder, evening workshop work) is actually one of the few legitimate reasons to have a larger battery bank because you genuinely need high-power energy after sunset. That’s very different from heating water, which is a low-grade, time-flexible load.
Also, your ROI recovery (~27% in 1.5 years) is solid but I’d argue that’s largely driven by:
1.Tariff escalation
2.High self-consumption
3.Avoided downtime
Not necessarily by how the batteries are being cycled overnight. So I think the middle ground is:
Oversize for independence if that’s your goal absolutely valid.
But once you’ve got that system, run it in a way that prioritises solar-first, battery-second, especially for thermal loads like geysers.
Different goals, same physics and the physics still favours using the sun directly wherever you can. 😜

I agree with you no problem discharging at 0.5C but if the op doesn't need it to be set higher than 120A as the installer set it up then it is fine why change it. If the load depicts higher draw from the batteries there is ample capacity.
I was merely responding to the op's question and referenced the spesification sheet.

You are spot on at 120A if you read the warrenty clause on the OEM spesifications. 43A x 3 = 129A charge discharge for the 3 batteries in parallel. See bottom section of spec sheet highlighted in yellow.