macafrican

If I had decent roof space at my house, I would go almost-off-grid:
council supply goes to a new small DB that has a plug for just in case next to it and the only other breaker is a battery charger that can be managed by data signal relay.    This council connection is 20A, so carries the smallest basic fee if any at all.    That gives you about 2kWdc charge capacity if ever needed.
entirely separate from council is the old DB running off battery inverter.    In effect the battery is pure isolation from grid.   It normally charges from solar input to battery inverter but if needed a relay allows the grid connected charger to also charge the battery.    All your house loads sit on this DB.
Because impossible to export, the NRS size limit on solar does no apply.
Obviously must still comply with electrical and building safety rules & regulations, but none of the other stuff.
Why not completely off-grid?    Most council bylaws allow them to charge you for there being a connection available, even if you don’t use it.   This way you get some benefit from the small connection - that 2kW charge can help a lot in bad solar times.
When I sell my house I will build a single floor home with the right roof to put up a LOT of solar.    My current house has awkward roofs and is too big for my 8kW of solar.

It’s really not that big an issue and we have NRS that limit capacity export.     The variability in a minute of my solar is LESS than the variability in a minute of my loads.     The grid handles fine my loads ranging eg between 200 and 250kw in a minute, it handles exact same way, my solar export ranging between 200 and 250 kw.
As long as users stick to decent equipment and connect correctly, distributed generation actually IMPROVES the distribution grid because we help lift voltage in the street.    When they get smart, we will get to stage where I only export in peak time of use when they have a big need but then they need to pay me properly.   Then I’ll export from battery a constant kVA value.
There are far too many people spreading doubt, uncertainty, fear.     Usually those councils that don’t like losing revenue, or people selling solutions that need to pretend things are difficult.

It’s really not that big an issue and we have NRS that limit capacity export.     The variability in a minute of my solar is LESS than the variability in a minute of my loads.     The grid handles fine my loads ranging eg between 200 and 250kw in a minute, it handles exact same way, my solar export ranging between 200 and 250 kw.
As long as users stick to decent equipment and connect correctly, distributed generation actually IMPROVES the distribution grid because we help lift voltage in the street.    When they get smart, we will get to stage where I only export in peak time of use when they have a big need but then they need to pay me properly.   Then I’ll export from battery a constant kVA value.
There are far too many people spreading doubt, uncertainty, fear.     Usually those councils that don’t like losing revenue, or people selling solutions that need to pretend things are difficult.

Steve, we started process in early 2023, the documentation then did not indicate 1250A breaker in 500kW bypass then (just like it incorrectly indicated not only the generator integration operation and the free space needed around isolation transformer but luckily my bunker could deal with that).   The RFP respondents were all large professional installers (over 200 professional electrical engineers including panel members on NRS) and all had to calculate and specify cables for distance between my LV room and my smartgrid bunker, for me to do the underground trunking in advance.

Coulomb : as I have it there is no power factor ‘downstream’ of the rectifier as it is DC.     I am not sure as what good/bad power factor the rectifier presents to the generator on the upstream side, thought it would be 1 from perspective of the gennie.     After the DC level, the power factor are a function of my loads and inverter and impact of solar.    The financially relevant at the 11kV meter before my transformer is nowadays very good because I switched my old PFC back on after PCS presented 0.6 PF at 11kV level.
If generator is AC coupled, I normally have to relay PFC equipment off.     Going generator:rectifier:DC would mean I can leave my PFC on while generator runs.
 

256Vac also looks suspect.
I have a very large different technology and the data does not always tie up so that even measured over example a week, grid export = Solar + Recharge - Discharge - Loads = Grid.   But mine are small % difference not your range.

At family holiday house we have gas hot water to some bathrooms and it is battle to get to a comfortable shower temp and have it stay there.    A plumber told me the problem is the showers have mixer type taps (swing for temp and pull for volume).    Do you have a bath with separate hot and cold taps (even if they combine where it runs into bath) and does it also have the problem?
Plumber said there is way to have equal hot and cold temp but then the cold supply at gas geyser must also be the cold supply to the point of use and there is a then a gadget that ensures pressure to gas geyser is equal to the cold water that carries on to point of use.
We’re considering chopping out the damned mixers and replacing with plain old two tap system.

R1458 just to be connected.      My town has charged high availability for ages, but they do charge lower kWh.    It is a clever strategy to try and compete against solar.     Eg at my factory we now pay R441/kVA for highest half hour kVA but 144c energy charges.      So on a 300kVA peak demand, the cost before one kWh is R132k per month….     The only way to beat that is to add storage to solar and have an inverter that caps your grid draw at say 100kVA.     That battery, ignoring any increased solar use, generates R1m a year in direct savings.    It also means solar runs 100% during loadshed.   I haven’t used diesel since September 2023  

I briefly looked at converting a LandCruiser FJ40/45.    It would not be possible to register for public road use, as the regulations have not been sorted.    Have a look at www.evconvert.co.za 

We do use the PCS for peak-demand-shaving.     The generator only becomes relevant when battery is too low, for example loadshed and bad solar.    Generator then must put back 300 kWh and help with carry loads to extent the capped grid and small amount of solar are insufficient.

3 phase 100A per phase.  I thought for those that understand the maths it would have been commonsense.

It is inevitable that end-user tariffs will change to reflect realities.    You’re naive if you expect that you can connect solar, over-generate in the day and draw at night and only pay the for the net kWh consumed in a month.    The grid is not your loss-less and free battery and I won’t subsidize your model.   There are idiots with solar and storage that after evening loadshedding run their loads from grid AND recharge their batteries at 1C, in effect presenting double the kVA load on the rest of the country.   Stuff that!
By FAR the bulk of power and energy is delivered under Time of Use tariffs whether that be to metro or council or large Eskom direct customers.    That involves a NMD fee per kVA, a monthly peak demand fee per kVA, peak/standard/offpeak kWh fees and half a dozen other minor charges.   A few small towns are not on Megaflex.
I’ve done large solar since 2013 and my town was a bit silly at the beginning, giving us 1:1 credits.   I say silly because that ignores that they have the cost of reticulation and maintenance.    That party ended and they went stupid the other direction by slashing kWh (to 115c/kWh) so that solar is compared to that not 275c/kWh and because they know there will be a half hour somewhere in the month that the sun don’t shine, they smacked R415/kVA peak demand.   My response was adding more solar and more storage and grid shaving because I’d be stupid not to.   So now have 425kW solar and 600kWh battery on one of my transformers, will do the others this year.
A fair system imposes a fixed fee based on the size of connection you want available (they reserve and pay for that NMD with Eskom), a variable fee based on that month actual peak kVA (they also pay Eskom for that) and then a Time of Use tariff for kWh.    The fastest way we get people to swap power demand out of peak periods (when the country is paying for R8/kWh peaking diesel and implementing loadshedding) is that peak period energy costs R5/kWh and off-peak costs R1/kWh.   Maybe in a year or two my council pays me to inject 200kVA in peak periods.   I can program that and then recharge between 11PM and 4AM.    If they deducted 100kVA from my demand fees for the 200kVA that I wheel to them plus they pay me half their Megaflex peak cost per kWh, they score and I score.
Smarter tariffs need not be bad for consumers, it can also open up benefits.

It's the other way around - @macafrican is saying that at the Eskom customers WITHOUT solar are currently paying on behalf of the people that DO have solar.
The old tariff structures were fine when nobody was using solar, but now more and more solar is buggering up the assumptions that are underlying the old billing.
The point is, the true cost is made up of a fixed rate for the infrastructure before you've even used a single unit, and a variable rate for burning coal to produce power. This is what is charged to your municipality by Eskom. However, the municipality in turn was always charging the individual households NO fixed charges, and everyone the same fee per unit, probably for simplicity. If everybody keeps buying from Eskom, the billing models make sense. But when Solar users are buying for example just 10 units per month, then they are not buying enough units anymore to cover the fixed charges of maintaining the grid. That cost then falls to the remaining people that are still buying around 600 or whatever units. That's the less bad scenario when at least folks are using batteries.
A further problem, on top of that, and that's the case here, there's the issue that it costs Eskom differently at different times of the day to generate power, low-cost during the night and most of the day when generating from coal, and a much higher cost during peak time when they have to run expensive gas and diesel peaker plants. Let's say - just making up the numbers for illustration but it's not far off - that it costs 80c per unit to produce power at noon, and R8,00 per unit to produce power at 6pm - as long as people are paying around R2.68-R3.71 all around the clock, it sort of works out. You "overpay" at noon, and you "underpay" at 6pm but on average it's fair. Now what happens with solar is that people are not paying anything during the day anymore, but at night they want to buy the most expensive power at below the cost of generation. The utility can only lose like that. Not just losing out on sales, but actively losing money that other customers must somehow cough up for. And the solar user is either oblivious, or doesn't care, and gets indignant about it, as if buying solar power gives him a right to do what he wants when he wants no matter the costs to others.
Just in principle, the only fair way is to start passing on the billing to the end user in the same way as the municipality gets billed by Eskom. A fixed rate, combined with a time-of-use billing, including also a time-appropriate rate for selling into the grid. At appropriate corruption-free price levels, whatever that may be. And it does change the economics around choosing the right solar system if you're buying new, and it screws up the assumptions you may have had if you've already bought.
 
 

It is inevitable that end-user tariffs will change to reflect realities.    You’re naive if you expect that you can connect solar, over-generate in the day and draw at night and only pay the for the net kWh consumed in a month.    The grid is not your loss-less and free battery and I won’t subsidize your model.   There are idiots with solar and storage that after evening loadshedding run their loads from grid AND recharge their batteries at 1C, in effect presenting double the kVA load on the rest of the country.   Stuff that!
By FAR the bulk of power and energy is delivered under Time of Use tariffs whether that be to metro or council or large Eskom direct customers.    That involves a NMD fee per kVA, a monthly peak demand fee per kVA, peak/standard/offpeak kWh fees and half a dozen other minor charges.   A few small towns are not on Megaflex.
I’ve done large solar since 2013 and my town was a bit silly at the beginning, giving us 1:1 credits.   I say silly because that ignores that they have the cost of reticulation and maintenance.    That party ended and they went stupid the other direction by slashing kWh (to 115c/kWh) so that solar is compared to that not 275c/kWh and because they know there will be a half hour somewhere in the month that the sun don’t shine, they smacked R415/kVA peak demand.   My response was adding more solar and more storage and grid shaving because I’d be stupid not to.   So now have 425kW solar and 600kWh battery on one of my transformers, will do the others this year.
A fair system imposes a fixed fee based on the size of connection you want available (they reserve and pay for that NMD with Eskom), a variable fee based on that month actual peak kVA (they also pay Eskom for that) and then a Time of Use tariff for kWh.    The fastest way we get people to swap power demand out of peak periods (when the country is paying for R8/kWh peaking diesel and implementing loadshedding) is that peak period energy costs R5/kWh and off-peak costs R1/kWh.   Maybe in a year or two my council pays me to inject 200kVA in peak periods.   I can program that and then recharge between 11PM and 4AM.    If they deducted 100kVA from my demand fees for the 200kVA that I wheel to them plus they pay me half their Megaflex peak cost per kWh, they score and I score.
Smarter tariffs need not be bad for consumers, it can also open up benefits.

It is inevitable that end-user tariffs will change to reflect realities.    You’re naive if you expect that you can connect solar, over-generate in the day and draw at night and only pay the for the net kWh consumed in a month.    The grid is not your loss-less and free battery and I won’t subsidize your model.   There are idiots with solar and storage that after evening loadshedding run their loads from grid AND recharge their batteries at 1C, in effect presenting double the kVA load on the rest of the country.   Stuff that!
By FAR the bulk of power and energy is delivered under Time of Use tariffs whether that be to metro or council or large Eskom direct customers.    That involves a NMD fee per kVA, a monthly peak demand fee per kVA, peak/standard/offpeak kWh fees and half a dozen other minor charges.   A few small towns are not on Megaflex.
I’ve done large solar since 2013 and my town was a bit silly at the beginning, giving us 1:1 credits.   I say silly because that ignores that they have the cost of reticulation and maintenance.    That party ended and they went stupid the other direction by slashing kWh (to 115c/kWh) so that solar is compared to that not 275c/kWh and because they know there will be a half hour somewhere in the month that the sun don’t shine, they smacked R415/kVA peak demand.   My response was adding more solar and more storage and grid shaving because I’d be stupid not to.   So now have 425kW solar and 600kWh battery on one of my transformers, will do the others this year.
A fair system imposes a fixed fee based on the size of connection you want available (they reserve and pay for that NMD with Eskom), a variable fee based on that month actual peak kVA (they also pay Eskom for that) and then a Time of Use tariff for kWh.    The fastest way we get people to swap power demand out of peak periods (when the country is paying for R8/kWh peaking diesel and implementing loadshedding) is that peak period energy costs R5/kWh and off-peak costs R1/kWh.   Maybe in a year or two my council pays me to inject 200kVA in peak periods.   I can program that and then recharge between 11PM and 4AM.    If they deducted 100kVA from my demand fees for the 200kVA that I wheel to them plus they pay me half their Megaflex peak cost per kWh, they score and I score.
Smarter tariffs need not be bad for consumers, it can also open up benefits.

There are places where that is in place by way of wheeling.    So Acme buys energy from SolaRus and basically council charges a toll fee for using its reticulation.   Putting in place new reticulation would be insanely inefficient.    Where that gets hairy is whether the wheeling is firm or not = whether council charges NMD and Demand fees on Acme’s full 2MW connection or the half that SolarUs is not providing.

It is inevitable that end-user tariffs will change to reflect realities.    You’re naive if you expect that you can connect solar, over-generate in the day and draw at night and only pay the for the net kWh consumed in a month.    The grid is not your loss-less and free battery and I won’t subsidize your model.   There are idiots with solar and storage that after evening loadshedding run their loads from grid AND recharge their batteries at 1C, in effect presenting double the kVA load on the rest of the country.   Stuff that!
By FAR the bulk of power and energy is delivered under Time of Use tariffs whether that be to metro or council or large Eskom direct customers.    That involves a NMD fee per kVA, a monthly peak demand fee per kVA, peak/standard/offpeak kWh fees and half a dozen other minor charges.   A few small towns are not on Megaflex.
I’ve done large solar since 2013 and my town was a bit silly at the beginning, giving us 1:1 credits.   I say silly because that ignores that they have the cost of reticulation and maintenance.    That party ended and they went stupid the other direction by slashing kWh (to 115c/kWh) so that solar is compared to that not 275c/kWh and because they know there will be a half hour somewhere in the month that the sun don’t shine, they smacked R415/kVA peak demand.   My response was adding more solar and more storage and grid shaving because I’d be stupid not to.   So now have 425kW solar and 600kWh battery on one of my transformers, will do the others this year.
A fair system imposes a fixed fee based on the size of connection you want available (they reserve and pay for that NMD with Eskom), a variable fee based on that month actual peak kVA (they also pay Eskom for that) and then a Time of Use tariff for kWh.    The fastest way we get people to swap power demand out of peak periods (when the country is paying for R8/kWh peaking diesel and implementing loadshedding) is that peak period energy costs R5/kWh and off-peak costs R1/kWh.   Maybe in a year or two my council pays me to inject 200kVA in peak periods.   I can program that and then recharge between 11PM and 4AM.    If they deducted 100kVA from my demand fees for the 200kVA that I wheel to them plus they pay me half their Megaflex peak cost per kWh, they score and I score.
Smarter tariffs need not be bad for consumers, it can also open up benefits.

OK. I'm out the country right now, and it seems I missed something? 1K for a connection? What sort of connection? 
Flip side of coin: 1K is about what residential users in Jhb are paying in fixed fees on the default tariff. So what's new? Also it is the City charging that, not Eskom. 
I'm on prepaid in Jhb. Currently no fixed fees, though that is about to change. And a comparison with my neighbour illustrates the problem. He pays ±1K per month just for being connected. I pay more per kWh but no flat fees. Yet the City must provide us both with the same infrastructure. 
Consumers will always play the tariffs game to get the lowest price. We're silly if we think the providers will not react to that. 

We had a lot of solar from early but that did not solve the problems of (1) solar throttle in loadshed (2) tariffs that target solar because council knows there will be THAT damned period of high loads, low solar.    My council in 2024/25 year is R508/kVA.    Hit 300 loads even once that month, over R150k hit.      The site generates about 75% of its consumption, so our kWh net from grid is small.
We asked 5 people to tender and most came up with Atess and variety of batteries.    We went 500kW Atess plus 600kWh IES battery bank.    500kW leaves headroom for growth in peak demand that is around 330kVA.
The kit sits in nice solid rooms with thick walls and concrete slab for roof.   Battery room can take about another 400kWh.    Another room has the PCS and isolation transformer.   Bypass Cabinet sits in LV Distribution room which was where grid and solar were.  Beware : the 500kW Atess bypass cabinet has 1250A breakers, so your cables to PCS+Transformer have to be big.
One business goal was seamless avoid generator, which is working well.   Have not needed gennie since end of September when we were running R80k a month in diesel.
The primary business goal was capping grid.    We now cap grid at under 100kW, which saves us say 225 kVA peak or over R100k a month.   If council reverts us to Time Of Use tariff, we can change rules to run off battery in peak tariff periods of day.
The project is big and complex and we had our challenges in terms of Promise meets Reality.     The only way to get acceptable power quality was reinstating the very large Power Factor Correction we had before smartgrid.    They said they’d handle it, but just looking at their capacitors versus what we had, is like a can of Red Bull compared to a Heineken keg.    Since then, power quality solved and PF measured at council MV level at full load is over .98.
We had to tap dance outside system to make solar (SolarEdge not Atess) work like it should and can.    If loadshed is after 10AM in sunny months, we don’t touch battery and in fact have to throttle solar if battery is over 95%
Today it rained all day so little solar even from the 425 on that network.    We kept grid cap in place and dropped battery to about 70% so despite nerves, it worked.
Last remaining tap dance is a rational way to integrate generator.     The Atess approach is pathetic, they may as well tell you to put an ATS between bypass cabinet and house loads.    Looking at two routes, will update how that goes.

We had a lot of solar from early but that did not solve the problems of (1) solar throttle in loadshed (2) tariffs that target solar because council knows there will be THAT damned period of high loads, low solar.    My council in 2024/25 year is R508/kVA.    Hit 300 loads even once that month, over R150k hit.      The site generates about 75% of its consumption, so our kWh net from grid is small.
We asked 5 people to tender and most came up with Atess and variety of batteries.    We went 500kW Atess plus 600kWh IES battery bank.    500kW leaves headroom for growth in peak demand that is around 330kVA.
The kit sits in nice solid rooms with thick walls and concrete slab for roof.   Battery room can take about another 400kWh.    Another room has the PCS and isolation transformer.   Bypass Cabinet sits in LV Distribution room which was where grid and solar were.  Beware : the 500kW Atess bypass cabinet has 1250A breakers, so your cables to PCS+Transformer have to be big.
One business goal was seamless avoid generator, which is working well.   Have not needed gennie since end of September when we were running R80k a month in diesel.
The primary business goal was capping grid.    We now cap grid at under 100kW, which saves us say 225 kVA peak or over R100k a month.   If council reverts us to Time Of Use tariff, we can change rules to run off battery in peak tariff periods of day.
The project is big and complex and we had our challenges in terms of Promise meets Reality.     The only way to get acceptable power quality was reinstating the very large Power Factor Correction we had before smartgrid.    They said they’d handle it, but just looking at their capacitors versus what we had, is like a can of Red Bull compared to a Heineken keg.    Since then, power quality solved and PF measured at council MV level at full load is over .98.
We had to tap dance outside system to make solar (SolarEdge not Atess) work like it should and can.    If loadshed is after 10AM in sunny months, we don’t touch battery and in fact have to throttle solar if battery is over 95%
Today it rained all day so little solar even from the 425 on that network.    We kept grid cap in place and dropped battery to about 70% so despite nerves, it worked.
Last remaining tap dance is a rational way to integrate generator.     The Atess approach is pathetic, they may as well tell you to put an ATS between bypass cabinet and house loads.    Looking at two routes, will update how that goes.

We had a lot of solar from early but that did not solve the problems of (1) solar throttle in loadshed (2) tariffs that target solar because council knows there will be THAT damned period of high loads, low solar.    My council in 2024/25 year is R508/kVA.    Hit 300 loads even once that month, over R150k hit.      The site generates about 75% of its consumption, so our kWh net from grid is small.
We asked 5 people to tender and most came up with Atess and variety of batteries.    We went 500kW Atess plus 600kWh IES battery bank.    500kW leaves headroom for growth in peak demand that is around 330kVA.
The kit sits in nice solid rooms with thick walls and concrete slab for roof.   Battery room can take about another 400kWh.    Another room has the PCS and isolation transformer.   Bypass Cabinet sits in LV Distribution room which was where grid and solar were.  Beware : the 500kW Atess bypass cabinet has 1250A breakers, so your cables to PCS+Transformer have to be big.
One business goal was seamless avoid generator, which is working well.   Have not needed gennie since end of September when we were running R80k a month in diesel.
The primary business goal was capping grid.    We now cap grid at under 100kW, which saves us say 225 kVA peak or over R100k a month.   If council reverts us to Time Of Use tariff, we can change rules to run off battery in peak tariff periods of day.
The project is big and complex and we had our challenges in terms of Promise meets Reality.     The only way to get acceptable power quality was reinstating the very large Power Factor Correction we had before smartgrid.    They said they’d handle it, but just looking at their capacitors versus what we had, is like a can of Red Bull compared to a Heineken keg.    Since then, power quality solved and PF measured at council MV level at full load is over .98.
We had to tap dance outside system to make solar (SolarEdge not Atess) work like it should and can.    If loadshed is after 10AM in sunny months, we don’t touch battery and in fact have to throttle solar if battery is over 95%
Today it rained all day so little solar even from the 425 on that network.    We kept grid cap in place and dropped battery to about 70% so despite nerves, it worked.
Last remaining tap dance is a rational way to integrate generator.     The Atess approach is pathetic, they may as well tell you to put an ATS between bypass cabinet and house loads.    Looking at two routes, will update how that goes.