Experience getting City of Cape Town to sign-off a zero feed-in system with inverter maximum output greater than 3.5kVA ?

[phil.g00]

@Fuenkli, Yes and No really, Yes you're right they had initially regulated for an all or nothing very intolerant frequency shift.

But, No even though the cliff is now a slope, they still all respond in unison and still bottom out at the end of that slop all together.

[Fuenkli]

1 minute ago, phil.g00 said:

No even though the cliff is now a slope, they still all respond in unison and still bottom out at the end of that slop all together.

I agree. But somehow they do not have the problem we think we might have with connecting a lot of PV capacity to the grid. I admit I am not familiar with the German requirements regarding maximum capacity of grid tied inverters. All I say is that other nations are way ahead of us in this field and we could probably implement their approach.

[phil.g00]

Yes, I think there needs to be a sea-change, whereby the domestic solar IPP is not seen as part of the problem, but as part of the solution.

That is what I believe the later German regulations were designed to do.

[___]

6 hours ago, phil.g00 said:

This is what i am putting forward, is a simple grid-owned MCB solution that is sized for the absolute current value of 25% as being an option.

If I may be a tad brand specific... On the Victron Inverters there is a setting called the AC Input limit. If you set that to 15A (for example), then it will only ever import 15A and also only ever export 15A. Never more. The feature is actually there from way back when you'd have your boat moored in a harbour with weak shore power. You could then limit the amount of power you use from shore power and take the rest from the batteries. This feature works in both directions on an ESS system.

If you combine that with an MCB, as suggested, then 1) you can't feed in more than 15A, and 2) if you do, say you go and change the setting on the CCGX, then the breaker trips.

This might mean you have to carefully put large loads on the grid side. On an ESS system, for example, if the battery drops too low then it cannot take the difference from the batteries, and if at that point you use more than 15A the inverter will go into passthru mode and the breaker will trip. But if properly designed, then it's not insurmountable.

[phil.g00]

13 minutes ago, plonkster said:

This might mean you have to carefully put large loads on the grid side.

I think we have to operate in the South African reality, that the authorities believe ( and probably with good reason), if the customer is in charge of overseeing himself that that is not going to be a reliable or acceptable solution..

So although brand specific solutions are available, I think that that MCB has to sit in their meter box, under their control, before this would be deemed an acceptable alternative to them.

But all it would it need is the same sign-off as anybody who installs these meters does anyway. 

A possible alternative solution that may be acceptable is that your 80A supply is split in their meter box into a 60A MCB and a 20A MCB, and solar generation is strictly only permitted on the 20A supply. That would need a COC sign-off to say it is, but still should alleviate the generation capacity limit concerns.

[Guest]

Or just get a inverter that matches the breakers amperage, as is, and be done with it.  

Not only won't you need to upgrade the breakers, or have 2 installed (although that is a clever idea), the whole cost comes down.

 

[phil.g00]

2 hours ago, The Terrible Triplett said:

Or just get a inverter that matches the breakers amperage, as is, and be done with it.  

Both are possible of course, but it's more practical if Mohamed goes to the mountain.

[cpbotha]

@Rautenk and others who are into this ( @plonkster , I have just run into a detail in a GoodWe installation manual (and then in the GoodWe docs) that sheds some more light on the "the storage can be used to reduce the effective size of the generator" question.

If you look at page 3 of http://alpspower.com.au/wp-content/uploads/2017/12/Goodwe-5048ES-with-BYD-B-Box1.pdf take note of the following text:

"Do NOT connect the grid to inverter's back up part under any
scenario. The inverter could be damaged if there is power flows
into the inverter from back up part.

To improve the system reliability, Goodwe recommends installer to
install a changeover switch between back up load and grid tied
load. Then for rare case, if the inverter is faulty, customer
could manual switch all essential house loads back to the grid
side."

In other words, all of the essential loads are in fact disconnected from the grid when the inverter is down. When the inverter trips (as defined in the NRS document) the essential loads CAN'T be transferred to the grid, only loads on the non-essential circuit.

So, if the inverter generates its maximum 5kVA and for argument's sake is not using any grid at that moment, and then trips / shuts down, only 5kVA minus 4.6kVA (its backup power) which comes to 0.4kVA of load can then be transferred to the grid. In other words, this specific setup will easily satisfy the 25% rule.

A demonstration of this would be to have e.g. a 1kVA aircon on the essential loads circuit, and then manually shut-off the inverter, at which point the aircon will simply die, and zero change in power flow will show on the grid side.

Does this make sense?

Edited February 12, 2019 by cpbotha

[cpbotha]

Just to add to that, below is a section from page 8 of the GoodWE EM series user manual ( http://www.goodwe.com/Public/Uploads/sersups/EM USER MANUAL.pdf ) where they recommend installing an SP3T switch to be able to manually switch backup loads back to the grid if required.

[___]

8 minutes ago, cpbotha said:

In other words, all of the essential loads are in fact disconnected from the grid when the inverter is down

Yeah... but they want to tell you HOW you are supposed to do this.

1. Presently, islanding is done by simply switching off the grid and leaving everything else on.

2. The way they want you to do it, is to first switch off everything, then safely create a dead island (if required), and then switch the island back on.

If you look at the CoCT diagrams, they seem to have the idea that there are two inverters, one that turns the DC into AC and feeds it into the grid, and a second one that feeds the battery power into the same grid... and the idea is that you can only turn on the second inverter after you have created an island. It does not seem to make provision for cases where one inverter does both things, that is to say, for the typical peak-shaving+backup setup.

If this is the case, it is a shortcoming in SANS in my opinion.

[___]

Let me consider again, for a second, the block diagram of what I think is a Goodwe:

 

Note the series disconnects 1 and 2.

1. When in bypass, 2 will be closed and 1 will be open.

2. When there is an outage, 1 will be closed and 2 will be open.

3. When running grid-parallel, both 1 and 2 will be closed.

The question is, what happens vs what should happen if the power fails while we're running grid-parallel.

As far as I know, all that happens is 2 is opened and we go on our merry way. This is apparently not good enough.

What you are supposed to do (if I understand all this correctly), is open both 1 and 2, then check that 2 really is open (using some sort of feedback mechanism), and then re-close 1.

Maybe this is already done... but it would have to be REALLY fast. 20ms or less.

[Guest]

I did not know this from GoodWe.

Have done all these loads off-grid with solar when Eskom goes off:

... dust cleaner as in a Kirby, 2kw kettle and a large 1400w jackhammer with the normal will always be loads on.

Some of it will be running when the grid comes back on. 

@plonkster please tell me we can do that?

[___]

5 hours ago, The Terrible Triplett said:

 please tell me we can do that?

I had an interesting result with my Bosch front-loader this morning. It was running fine from the MultiGrid (I just set it to cold wash to avoid the heating cycle). Then my wife started the hair dryer (which added about 2kw to the already 800W or so we were running, so close to capacity). Suddenly the washing machine started to struggle. It has a different kind of motor (not quite like the usual induction stuff) and it electronically ramps it up... so I could hear the motor struggling to start moving... and then the machine stopped. It looks as if the heavy load caused the voltage to sag as the machine tried to start up the motor, and it gave up.

So can it run it? Definitely yes. But you do see some weirdness at the edges. I've run my non-inverter ACs from it as well, but only up to 12000 BTU.

Edit: The blue boxes are low-frequency designs with large transformers. They tend to be better at this sort of thing.

Edited February 12, 2019 by plonkster

[cpbotha]

@plonkster I have found the document from which your block diagram was taken. It's the TÜV Rheinland certificate of conformity to NRS-097-2-1 2017 for the GoodWe EM series, and it's quite insightful to read through.

You can find a PDF at https://solaradvice.co.za/wp-content/uploads/2017/10/EM-SA-test-report.pdf

(I could not attach the 3.5MB PDF due to the 3MB attachment limit here.)

In the document, you can find how the certification authority directly handles each of the NRS requirements where it concerns the inverter being investigated, for example:

NRS clause 4.2.2.2.1 "The embedded generator shall be equipped with a disconnection device, which separates the embedded generator from the grid due to abnormal conditions" ... with the response "PV inverter provided two relays in series used in each line and neutral as disconnection devices" and verdict "P" for pass. That easy.

You'll probably find section 4.2.2.4 on anti-islanding interesting.

 

(With regard to my own plans: I have decided that it's going to take too long for CoCT to catch up with us, and am going to bite the bullet with the GoodWe EM 3.6. It's on the 2017 list, and it satisfies the max power output requirement.

In other more interesting news: The director of SARETEC has told me that CoCT is flying in an expert from Germany who will work for them for the coming two years, also to help with the applied regulatory aspects of PV Solar SSEG. I also suggested to him that CoCT should have a workshop with a number of the most knowledgable and trained folks from this forum, to help disseminate CoCT regulatory information. Watch this space...)

 

[Guest]

@Rautenk and I would be very interested in this ... 

18 hours ago, cpbotha said:

In other more interesting news: The director of SARETEC has told me that CoCT is flying in an expert from Germany who will work for them for the coming two years, also to help with the applied regulatory aspects of PV Solar SSEG. I also suggested to him that CoCT should have a workshop with a number of the most knowledgable and trained folks from this forum, to help disseminate CoCT regulatory information. Watch this space...)

... seeing that my application can now go in seeing as the CoC is done and dusted AND the engineering has been, the only contentious issues being, seeing as the inverter was tested onsite: 

 

 

[Elbow]

5 hours ago, The Terrible Triplett said:

@Rautenk and I would be very interested in this ... 

... seeing that my application can now go in seeing as the CoC is done and dusted AND the engineering has been, the only contentious issues being, seeing as the inverter was tested onsite: 

 

 

 

So I think its very interesting what @Rautenk wrote in your application.  Since this is exactly the area where its not clear what is required and why.

Unfortunately I'm sure it will be months before there is any clarity.

The fun one is that you can't be compliant to both the NRS and the CoCT in respect of voltage levels?  Seems like your application also claims that CoCT is not compliant to their NERSA licence?

Or am I reading between the lines wrongly?

 

 

[Rautenk]

@Elbow, best I clarify before any misunderstanding! 

 

CoCT requests that you use NRS as your settings levels, they mention it quite a bit and so you have to use it. The inverters are also tested to it and the settings are loaded accordingly.

What the Grid Code require is different. City of Cape Town can request anything in terms of Embedded Generation as long as the Combined Effect is as per the SA Grid Codes at their point of connection to the Eskom Grid. So it is their responsibility to model their network and ensure that all things combined leads to compliance.

Then also, Grid Code needs to change (again), it is just a long process... So by the time that the penetration levels in CoCT grid is high enough to affect their "inertia" the grid code and NRS should most probably be aligned. 

 

When I have some spare time I will go back through all of the comments and see if there is anything else to answer for... Otherwise send me a Private Message and I can attend to it.

 

 

[Elbow]

1 hour ago, Rautenk said:

What the Grid Code require is different. City of Cape Town can request anything in terms of Embedded Generation as long as the Combined Effect is as per the SA Grid Codes at their point of connection to the Eskom Grid. So it is their responsibility to model their network and ensure that all things combined leads to compliance

Ah- understand what you are saying and the distinction.

Swaar-stroom. It’s a learning curve.

 

[Guest]

14 hours ago, Elbow said:

Swaar-stroom

Ja, like paddling up the creek without a paddle.