Fuenkli

I agree this is another option but what about our carbon footprint? I think @plonksteronce calculated that charging an EV with dirty Eskom power actually increases your CO2 emissions compared to driving an internal combustion engine car 🤣. 

I agree. If the inverter has the capability to communicate with the specific BMS you obviously want to make use of it. But there is a lot of inverter/Battery combination out there who can not communicate with each other at the moment. But (based on my experience) If the inverter has accurate SOC calculation and is programmed in line with the appropriate battery charge and discharge parameters the system will work well even without a direct communication link between the inverter and the battery.       

that is incorrect. The diodes will bypass the shaded panel/s and the string will now operate at a lower voltage. The output from the unshaded panels is not affected. The problem is, that without a good shadow management the inverter will not find the correct mpp and as a result operate the string at the wrong voltage. Another problem is, that the voltage of the string might fall below the inverter minimum operating threshold. This is why long strings (if some shading occurs) are better than short ones.  

that is incorrect. The diodes will bypass the shaded panel/s and the string will now operate at a lower voltage. The output from the unshaded panels is not affected. The problem is, that without a good shadow management the inverter will not find the correct mpp and as a result operate the string at the wrong voltage. Another problem is, that the voltage of the string might fall below the inverter minimum operating threshold. This is why long strings (if some shading occurs) are better than short ones.  

You are currently making good return and your system will have a very high IRR, but it seems you are exporting more than what the CoCT SSEG system allows for (ie you are not currently a net consumer). Net consumer is calculated on annual consumption and production so you might still be ok once winter is taken into consideration, but seeing your impressive tracker results, I'm not sure
You are missing my point about having the option to add batteries, of course it will be more expensive and thus IRR will be potentially reduced but batteries are all about having power when Eskom can't supply - for most people (not you), eliminating loadshedding, is as important as the ROI. On a small residential system the Hybrid inverter may add R10K up-front, but then you are at least able to add batteries at a later point if the feed-in tariff is reduced (or eliminated) - or loadshedding ramps up even more.
One of the main problems with CoCT SSEG feed-in system is that you are effectively entering into a 1 year contract with the City on your 20-25 year solar investment. The monthly SSEG connection fee, tariff and feed-in tariff are changed annually which means your system IRR is very vulnerable, especially if it is dependent on the feed-in, to achieve savings (eg. City may only offer R0.33/kWh feed-in next year or R0.10/kWh or none at all.....). The R0.25/kWh incentive is for first year only and may not even apply to systems added in future years. If CoCT would offer a feed-in deal fixed for 5 years then we'd be able to do our sums and make much more informed decisions.
With a self-consumption battery system you will fix your R/kWh rate for life of system (currently R1 - R1.50/kWh) so it's a bit easier to work out the IRR as the Eskom kWh tariff historically has only ever gone up. Feed-in tariffs world-wide have mostly been coming down.
Currently, hybrid battery system in Cape Town expected pay-back is 5-8 years, grid-tie SSEG can be as low as 3-4 years (but no grid - no worky).

YEs, but that that should be big farms that are on the supply side of the grid and which whoever is controlling the grid can manage. Off shore wind farms for example. Solar farms in the Karoo.

that is incorrect. The maximum output power of the inverter is 4.6kW (essential and non-essential combined).

As others have mentioned, the starting point would be to have a look at the SEMS data. It'll give you an idea what the total load on the inverter is. There are physical limits to what the inverter can supply and going in blindly, without thorough understanding of your loads, is a recipe for disaster. Without data, it's like looking for a needle in a haystack.
Based on some comments above, it seems you only have the inverter tripping when the grid falls away and under load? That does sound like you battery is tripping, not the inverter. Again, as others have mentioned, one US3000 isn't really optimal for the inverter. You are technically limited to 1.7kWh worth of backup capability by the Pylontech battery, which is a 1/3rd of what the inverter is capable of. If your loads never exceed that, then all is good. But based on what you mention above your loads can easily surpass that as 1.7kWh is not a lot for a typical household. The fact that your inverter is now running in Backup mode and not tripping, further confirms that your battery pack is under provisioned for your loads.
The first thing to do is to have a look at the daily SEMS data to get an idea what's going on (please confirm that you have access to SEMS?). Once there's a day or two's worth of graphs then we can look at you switching off the grid and seeing what's drawing power. Unfortunately it's a little difficult when we don't know what is wired to the backup side of the inverter. I have a strong feeling in my gut we are going to circle back to insufficient battery peak delivery.

I don't think I know anybody who had a good 2020, but I do hope there are better things ahead for 2021. I have enjoyed this forum this year, and have learned from the helpful, knowledgeable members. 

I hope next year is better for y'all. At least you will be able to keep the lights on.

https://www.theguardian.com/environment/2020/nov/10/renewable-energy-covid-19-record-growth-2020

The switch with feedback to the inverter is the internal relay or contactor that the inverter uses to connect the load to the grid when in standby mode.
(This is only applicable for Passive standby UPS used as off-grid SSEG)
The inverter must know the state of this switch (not just based on what it wants the switch to do, because switches can become stuck or welded closed) so that it can never try to power the load from the inverter while the load is powered by the grid.
The Hager changeover switch is just a manual external changeover switch (It is good to have this in your drawing when submitting your application as the people approving them don't know the difference, they just look for a changeover so make sure to label it as such in the drawing)

There is no such things as AC or DC cable.  Many cables are of course designed with a particular purpose in mind, and are then labelled/sold as such.
Copper cable has 2 important basic properties:
The cross-sectional area determines the current carrying capacity.  You choose it (at the economic level) by the voltage drop (power loss) you are prepared to tolerate and (at the safety/legal level) by the cable's ability to shed the excess heat generated due to that power loss.  Note that this heat-shedding is dependant on the cable's installation: a cable in free air can disperse far more heat than one running in a conduit, especially if surrounded by other heat-producing cables. The voltage rating of the insulation.  A key difference between AC and DC is that 230V AC actually has a peak voltage of 325V There are other properties that are relevant for PV cables: they need insulation that is more resistant to UV than cables designed for indoor use.  PV cables also need very high insulation resistance (>1000V) if used in high voltage strings.
The important things is: if your peak voltage is below the cable's rated insulation voltage, and your (RMS) current meets the criteria above, it does not matter if you use push AC or DC through it.
 

@Richard Mackay,
The answer depends on your usage and assumptions.  For daily cycle (off-grid) use LiFePo4 batteries are much cheaper per kWh over the life of the battery.
Example: US3000: 6000 cycles x 80% DOD x 3.5kWh = 16800kWh in life.  At R20k per battery, this works out to about R1.20 per kWh.
Any reasonable lead acid battery will cost at least R2000 for 100Ah@12V.  Assuming you get 1000 cycles at 50% DOD total production is 1000 cycles x 50% DOD * 1.2kWh = 600kWh.  At R2000 per battery, this works out to R3.30/kWh
Of course it ALL depends on you assumptions.  Can you really get 6000 cycles from your Pylon? On the other hand, can you really get 1000 cycles at 50% DOD from lead-acid?  Perhaps you can half Li-ion cost by getting local "second-life" batteries?
Other factors: Li-ion takes up less space, need less maintenance, has BMS and balancers built in, has higher round-trip efficiency and hence generates less heat, is cooler😎
On the other hand, for UPS/standby (anti-load shedding) use lead-acid is a good option.  Certainly you will never get near using 6000 cycles in that scenario, so Li-ion's biggest advantage is negated.
I bought Pylons - if I was doing it again I would look very hard at the "second-life" option.  My batteries only go through about 120 full cycle equivalents per year, so I will never get near using 6000 cycles.  It may be a bit of overkill.

this is the problem with an island system. It has to cater for the maximum power consumption of your house. My advice is to rather go grid tied. You can size the inverter for the average consumption and the grid will absorb the peaks.

Well in my case after they installed the meter my own accounting differed hugely with theirs (in their favor).
I did read the meter registers in the kiosk and found the meter reading to agree with the SAP system. However a short look further I noticed that they reversed one of the phases. That's of course not a good idea.
I took photos and sent them to COCT electricity east explaining the issue. Never got a reply. Eventually had enough of that and broke the seal and fixed the issue myself. Probably would have sent a tamper alarm but I took photos of the before and after (lots of them just in case).
Since then my own accounting and theirs match closely.
BTW with a three phase system at 100A you are only allowed a total of about 17KVA feed in - not 10KVA per phase. Your capacity can be higher but you have to prove that your system will never push more than the authorized limit.
I actually have a three phase 150A feed but had to downgrade it to a 100A feed as otherwise they wanted to force me onto a commercial tariff. That would have completely screwed me over...

@smurfdbn I am building a LiFePO4 battery with 16 of 2 cells from @Bain Viljoenin parallel. Got them from @Mike who resides in our village.
@Ian said In an earlier post that he did that and also put blocks of 3 cells in parallel for a neighbour. Perhaps he would be so kind and post his experience thus far for us to learn. As I understand it, parallel cells will self balance over time and the ANT BMS which I have should balance the pairs of cells. Multiple batteries, each with their own BMS have the problem of keeping the batteries in balance. If they are not, one or more with lower voltage may cause the ones with higher voltage to overcharge and then self-protect by disconnecting. There is after all no electronics keeping the banks in balance so no real difference to parallel cells. If a block of parallel cells have too low a voltage, it will cause some others to reach overcharge voltage and the ANT will detect that and also stop charging. 
Seems to me the parallel cell option may be preferable on balance. Will Prowse also has a video about it on YouTube. 
 

Well just received COCT bill for the month (up to 16th of October) - about the middle split for my annual estimation.

As you can see I am feeding in more than three times what I am taking at night. Obviously I am producing more than what I am feeding in as the property consumes power during the day - the feed in being the excess.
This compares well with my estimates - I should be able to get relatively close to net-zero as averaged over a solar year but probably will not quite get there unless I add a few more panels. But since I already have 42 panels on the roof it is getting a tad crowded so I would have to align additional panels facing west - a bit less efficient but not too much.
Considering this is the maximum size COCT allows for a residential installation and tariff you can see their calculations just allow net zero for this - anything smaller and its not worth your while. So the notion that COCT does not really want you to do this is a fair conclusion to make.
It took a year of fighting to get this signed off and in total it cost around half a million bucks. To be fair that includes a large battery bank and related inverters as well. Total installed inverter capacity is 30KW.

https://en.wikipedia.org/wiki/Basslink#:~:text=The Basslink electricity interconnector is,mainland to the George Town

I'm sure South Australia already has interconnects with neighbouring states Victoria and New South Wales, but I believe that they are upgrading some of those. The whole east coast of Australia (down to the island of Tasmania) is interconnected, and quite a lot of power gets moved around between states. Western Australia and the Northern Territory have their own separate grids (Australia is a big country, about the same area as the contiguous USA but with 6% of the population).
South Australia closed down all their coal fired generating plants years ago. They have some gas generators, but apart from that, they rely on interconnects with other states to supply any shortfall.

There has been zero communication from COCT. We will have to wait and see.
You are permitted to feed in as much as you like - subject only to the restriction on maximum power you may feed in based on size of your connection.
The limitation is monetary - so if you somehow manage to get to zero Rands as averaged over one year - you would have fed in three to four times what you actually took from the grid - slightly dependent on the size of your connection - the fixed daily charge is the same regardless of the size of your connection. If you have a single phase 65 amp feed - just forget the whole thing and just use your power yourself or start looking at using batteries.
This whole notion of using the grid as your battery - Hah ! Yeah right - it must be the Worlds most ineffective battery. You can do it if you feel your excess power is going to a "good home"...

This article explains the current state of play in the City
https://www.dailymaverick.co.za/article/2020-10-21-joburg-electricity-costs-spiral-as-city-suffers-covid-19-revenue-loss/

 
 

No one rushed to answer this one (and me being suspicious) I posted this on another local forum and the Terrible Tripplett responed:
Does this mean that even if you are a registered SSEG CoCT cannot buy your excess power due to the national contact they have with Eskom ?? 
Correct. Until the Constitution is changed, or it is gazetted, no-one but Eksom is allowed to sell power to anyone. If you are grid-tied, and do go the feedback route, you have to netmeter over 1 year, you owe them some, they owe you nothing.

That is about to change ... how, we will have to see.

My suggestions, and I've emailed them:
1) Bidirectional meter should be free, not costing me R12k+, offset that cost by paying me a lower rated per kw.
2) If I buy the birectional meter, pay me, ex VAT, Eskom rate per kw.
3) Daily connection fee of wot, +-R14 ex vat per day, nada, no, forget it. Keep the R150 ex VAT connection fee to help offset the ongoing grid maintenance costs.

Can only hope that:
1) They do consider the above for there are a lot of homes with spare, who can assist at no cost nor risk, to CoCT.
2) We, the small guys, can make a cent or two ...

Thing is though, on a 63amp breaker, you are limited to 3.5kw.
But if you can power your house, send the spare back to them, and make R10 per day ... that could be cool.

The proposed tariff for 2020/1, for residential properties who are net consumers and who sell "spare" power of less than 1 MW/h per month is 57.34 c. Plus you can't be on prepaid, you must be on a TOU tariff, your equipment must meet COJ standards (which implies an inspection) and you will need a bidirectional meter to be installed at your expense.

If you're on pre-paid then switching to TOU means about R600 of service fees each month.  COJ have been trying to force through a flat rate component for pre-paid connections. Two years ago it was removed because, the then Mayor said, somebody had sneaked it in after council had signed off on the new tariffs structure (at least that's what he said). More recently it was properly proposed then removed as part of relief under the pandemic, but it's not off the table. It will be less than the post-paid and TOU customers pay (as it should be. Pre-paid users pay more per kw/h). The long term objective seems to be to eliminate the gap between pre-paid and post-paid, which seems silly to me. If I pay in advance then that's good for the city's cash flow and they should give me some sort of consideration for that.

Richard, you're right, and you're wrong...

Some of us saw this trainwreck coming a long time ago and started working on our own solutions. Becoming power (or water) independent is a long and gradual process to which you must devote decades.

We also may or may not have had a long history of trying to sway official policy and direction, to little effect.

If you wait for policy to shift you should get used to cooking over an open fire and going to bed when the sun goes down.

Some good news.  It's taking a bit of time but I have my first data from the Goodwe coming into OH!