Stanley

As far as I know (I'm sure somebody will correct me if I'm wrong) this is only true if you are selling your house. i.e. When selling, you need a valid COC less than 2 years old. So if the COC was done more than 2 years ago, you will have to do it again. This is of course unless they have decided to change the rules.

Those old Powerstars communicate using Modbus RTU over an RS232 connection. I have attached the Modbus comms manual for you, which describes all the registers etc. although if MLT Coms is unable to get data or event logs, then my guess would be that the SD card which it uses to store all that data may have failed. Powerstar-Modbus Manual v2.pdf

This makes implementing a PV solution on the DC side much more complicated. I would definitely go for the AC route. The only disadvantage being the loss of production during a power failure.

I have found that pure grid-tied inverters (not hybrid inverters) are actually cheaper than charge controllers. Perhaps look at the Solis range of Inverters, they are very good and reasonably priced. I agree that you will lose a little bit of efficiency, but not so much that I would be concerned.

Why not just get a 3 phase grid-tied inverter and feed into the same 3 phase supply that is used for the rectifiers? That way you can supplement the energy used by the rectifiers without needing to change anything on them or interfere with them in any way.

There is a thread on this forum about it. Eskom are not planning on charging you more if you have solar panels. They want to change the entire tariff structure which, if implemented they way they want, would make owning solar panels less attractive. They want to introduce a base fee, irrespective of usage and then flatten the inclined block structure tariff so you don't pay much more for units above 350kWh or something. So you won't have a huge benefit anymore of shaving off a few kWh to drop you to a lower tariff block.

Another sneaky one that could get you is lighting. If you don't already have LED lights everywhere, then switching to LED lights could reduce your nighttime consumption significantly. When we moved into our current home, the previous owners had put in 100W halogen lights all over for some reason. The first thing I did was replace them with LEDs.

An overload is a load that is greater than the inverter's rating. Depending on how much the inverter is overloaded it may be able to handle it for a while. A short-circuit on the other hand is a bit more difficult to define, and is just a very low impedance. A short circuit (much like an overload) depends on what is supplying the power. You would agree that a 10kW load would be an overload for a 5kW inverter, but not for a 10kW inverter. Similarly a 20kW load might be so much of an overload for a 5kW inverter that it's over-current protection kicks in to protect it from blowing up. This would essentially be seen as a short circuit to the 5kW inverter, while a 10kW inverter would just see it as an overload. Now in the case of a power failure, we are talking about a load of hundreds of kW or even MW. Let's say for the sake of this discussion that your inverter is trying to power a load of 100kW on one phase of the supply. The impedance of that would be (230^2 / 100000) = 0.529 Ohms (Since P = V^2 / R so R = V^2 / P) Now I don't know about you, but I would regard an impedance of 0.529 Ohms to be pretty much a short circuit, but again that is subjective because a 100kW inverter would be happy supplying that load. All I was trying to do earlier was describe what happens during a power failure, and why certain things happen. i.e. Why the inverter output collapses to 0V before it can disconnect from the grid. I don't think any of us have inverters capable of supplying hundreds of kW to keep the grid up during a power failure. Although it could become a problem if enough people in a small area have inverters, as they could collectively keep the grid going which is one of the reasons why the anti-islanding protection in inverters has to be very good.

Not to a small inverter of only a few kW

Not necessarily. According to the CoCT rules, a passive standby UPS may be registered as off-grid embedded generation. A passive standby UPS is one that is not inverting while grid connected. i.e. It just passes the grid through to the load (it may however include a battery charger). As far as the inverter is concerned, the grid may as well be a short circuit when it is off. If the inverter is trying to produce 230Vac onto the grid, it will be trying to power the whole local grid. i.e. Transfromers and all the other users connected. This could be hundreds of kW or even MW of load.

No, a small change in frequency wouldn't make the dimmers switch on or off. It may cause the brightness to change briefly though. When the grid fails, it takes time for the inverter to realize that the grid is gone and time for it's contactor / relay etc. to open and disconnect it from the grid. In that time. the grid voltage has dropped to 0 very quickly, this is the first transient. Then the inverter needs to take over, ramping up it's voltage very quickly which would be the 2nd transient. Note that this is different from just turning off the circuit breaker supplying the inverter because during an actual power failure (i.e. Load shedding etc.) the supply to your entire neighborhood has been disconnected, so your inverter which is connected to the grid is essentially trying to supply the whole neighborhood for a few mS. So this appears as a short circuit very briefly until the contactor or relay can open to disconnect it from the grid. During that time the load will see the voltage drop to 0 and then climb very quickly again. This dip is usually about 20mS, but may be a bit more or less depending on the inverter and how it detects a grid loss and also what it uses as it's disconnecting device. For off-grid inverters where the inverter isn't actually inverting while the grid is present you will see something similar, but this will happen even if you just turn off the breaker supplying the inverter and not only during a real power outage. This is because the inverter also needs to detect the grid loss then disconnect from the grid and then start inverting. (A true hybrid inverter is always inverting even while the grid is present, so the problem is only that the grid becomes a short circuit when the power fails)

Switching on or off is caused by high frequency noise or transients. When load shedding starts or stops, it will be transients causing the switching on or off. I am not sure what solution there is for this problem. I would suggest asking the dimmer manufacturer if they know of anything.

Dimmers can be affected in a couple of ways. Most dimmers look for the zero-crossings of the AC signal and use that for their timing. Then depending on if they are leading edge or trailing edge dimmers, they either wait some time after the zero-crossing before turning on or they turn on at the zero-crossing and then wait some time before turning off. Either way, distortion of the AC waveform can cause the zero-crossing detection to trigger a little earlier or later on different cycles, making the brightness fluctuate. So if the brightness was constantly changing then it is most likely caused by some distortion when the hair dryer was on. Another possibility with bell-press dimmers (the ones that you turn on and off by pressing the same button that you hold to change the brightness) is that they have a small capacitor between live and the button input and high frequency noise can go through that capacitor making the dimmer think that the button has been pressed briefly. This will cause the dimmer to turn on and off. So if it was turning on and off then it is most likely caused by high frequency noise. Hair dryers are particularly nasty if you use them on their half power or half speed setting, because most hair dryers use a diode rectifier for the half speed or half power setting, which means they only draw current for one half of the AC wave. This causes transformers to saturate and then the waveform can get quite badly distorted.

Off-grid installations have always been allowed. There was some confusion around the rules, but that has been resolved quite a while ago now. The latest documentation from the CoCT is pretty clear about the requirements.

No. Most large PV installations have no battery storage, so they feed all the available PV power into the grid. Which means they cause a big dip in the demand from the utility centered around midday. They do little to nothing during the morning and evening peaks. This is what the utilities don't like. They have pretty much the same peak loads as before, but the base load becomes very small. This causes problems as they have to ramp generation up and down much more quickly and by a greater margin. In fact this is one time that I actually agree with eskom. If you have a PV system and are using the grid as storage, i.e. Feeding back excess power during the day and then buying it back at night so that your average for the month is basically 0, you have still used the grid and there are costs involved for the utility to provide you with that connection, so I agree that there should be a base connection fee to cover the fixed costs. The price per kWh can then come down.

This is only for grid-tied PV systems. For off-grid systems a COC is enough.

Those are rather expensive, but I agree he needs a battery inverter and batteries. I would recommend one of the MLT products. The price list can be found here: https://www.mltinverters.com/pricelist.html Since the OP has AC coupled PV already, the Oasis would probably work well as a UPS that supports having grid-tied inverters on it's output. He must just choose one that is of equal size or greater than his PV inverter. The Oasis 10H is quite a lot cheaper than the SMA 8.0

Unfortunately that is not possible without batteries. You could use a backup generator and the let the PV inverter connect to the generator output but then you need to control the PV production based on the generator load to make sure that you never feed back into the generator because that can damage it.

In grid-tied inverter terminology, islanding is a bad thing. If the grid fails, but the inverter doesn't detect the grid loss and continues to energize the lines on it's AC input, then it is said to be islanding (or creating an island). This is not allowed, and so inverter manufacturers have to employ several different methods of detecting this situation so that they can disconnect from the grid and prevent the creation of an island.

I used a small inline fuse on my pilot light for protection and I got a COC for my installation.

If the inverters are connected in parallel, shouldn't the generator be connected to both inverters?

Perhaps somebody who knows ICC well can answer this question then: If ICC is communicating with 2 batteries and 1 reports 100% SOC and the other 0%, will ICC report 50% SOC (i.e. The average of the 2)? A loose connection could definitely cause low voltage issues, let's hope that's all it is and the fix is as simple as a lock washer. Good luck

That battery voltage curve is typical of a lithium battery being completely discharged. i.e. The voltage remains pretty flat until the battery is close to empty and then the voltage starts dropping rapidly. This should not happen at 50%SOC. How are you reading SOC? i.e. Do you have comms to the battery itself or are you using a 3rd party device like a Victron BMV to measure energy in and out? Do you have more than 1 battery in parallel? If you have 2 for example, then maybe there is a bad connection on 1 of them so you are only using 1 of the batteries and therefore only have half the capacity available (hence being empty at 50%)

I have not used ICC, so I'm not certain how this works, but if it can read SOC from the battery and use that to control the inverter then yes, that sounds about right.

It is practically impossible to assign an SOC value to a particular battery voltage. The actual voltage will depend heavily on the load. At a very light load, 48V could be a very low SOC, while under a heavy load 48V could be a fairly high SOC. So unfortunately if your inverter doesn't support CAN comms with the Pylontech battery then you have to play it safe and use the recommended 48V otherwise you could have a situation where the battery shuts down on low SOC if the load is very light.