Hi all. 

I would appreciate any help with an problem I just recently noticed with my system. 

My setup:

1. MUST 3K 24v PH18-PLUS (Hybrid, grid feedback capable)

2. 5kwh 24v battery.

3. 1.2kwp of PV

4. AC connections: 

Main DB MCB -> Sub panel/DB MCB -> Inverter AC IN.

Sub panel/DB:

Inverter AC OUT -> MCB -> RCD -> Dedicated critical loads circuit.

I also have a Normally Closed contactor/relay, that does the N-G bond on the output, when grid is down (or MCB to inverter is OFF), since my inverter doesn't do N-G bonding automatically when in Solar/Battery mode.

Everything works fine, no nuisance tripping in any of the inverter modes (SUB/SBU). But up untill now, I have been using it with grid feedback disabled.

After enabling grid feedback, I noticed this problem:

If I turn off the input MCB (going to AC IN of the inverter), simulating a power outage, it causes the output RCD to trip immediately! This only happens when actively backfeeding. If I disable grid feedback, I can cutoff power to the inverter and everything continues to work fine.

After it trips, I can re-enable the RCD and it doesn't trip. I have tested this multiple times. 

I thought it might have something to do with double N-G bond on the output, but it shouldn't happen: When grid power is down (or I turn off the MCB going to AC IN), the NC contactor is de-energized and closes, making the bond.

Meanwhile the inverter's internal AC IN relay should open (I hear the click), so I shouldn't have another N-G bond from utility. Even so, it doesn't explain why this happens only when backfeeding.

Maybe when grid feedback is enabled, it takes more time for the internal AC relay to open? Does this make sense?

I gave this N-G bonding solution a great deal of thought and research on the forums, before deciding to go this way.

I didn't like the NEUTRAL to NEUTRAL bond solution, as it is controversial.

Any ideas? 

Thanks.

Edited June 14, 20241 yr by meetyg

1 hour ago, meetyg said:

Hi all. 

I would appreciate any help with an problem I just recently noticed with my system. 

My setup:

1. MUST 3K 24v PH18-PLUS (Hybrid, grid feedback capable)

2. 5kwh 24v battery.

3. 1.2kwp of PV

4. AC connections: 

Main DB MCB -> Sub panel/DB MCB -> Inverter AC IN.

Sub panel/DB:

Inverter AC OUT -> MCB -> RCD -> Dedicated critical loads circuit.

I also have a Normally Closed contactor/relay, that does the N-G bond on the output, when grid is down (or MCB to inverter is OFF), since my inverter doesn't do N-G bonding automatically when in Solar/Battery mode.

Everything works fine, no nuisance tripping in any of the inverter modes (SUB/SBU). But up untill now, I have been using it with grid feedback disabled.

After enabling grid feedback, I noticed this problem:

If I turn off the input MCB (going to AC IN of the inverter), simulating a power outage, it causes the output RCD to trip immediately! This only happens when actively backfeeding. If I disable grid feedback, I can cutoff power to the inverter and everything continues to work fine.

After it trips, I can re-enable the RCD and it doesn't trip. I have tested this multiple times. 

I thought it might have something to do with double N-G bond on the output, but it shouldn't happen: When grid power is down (or I turn off the MCB going to AC IN), the NC contactor is de-energized and closes, making the bond.

Meanwhile the inverter's internal AC IN relay should open (I hear the click), so I shouldn't have another N-G bond from utility. Even so, it doesn't explain why this happens only when backfeeding.

Maybe when grid feedback is enabled, it takes more time for the internal AC relay to open? Does this make sense?

I gave this N-G bonding solution a great deal of thought and research on the forums, before deciding to go this way.

I didn't like the NEUTRAL to NEUTRAL bond solution, as it is controversial.

Any ideas? 

Thanks.

You mention when you are feeding back when you switch the AC in off. Now the inverter should not be able to feed back as feeding back has to stop when there is no grid and it should actually feed back via the AC in which is connected to the grid side. 

It seems like the inverter is using PV to feed back via the back up side. 

The PV should only feed the back up load and battery charge when the grid is off via the AC in that you "trip" 

This is how I understand what is mentioned in your post. 

2 hours ago, meetyg said:

Hi all. 

I would appreciate any help with an problem I just recently noticed with my system. 

My setup:

1. MUST 3K 24v PH18-PLUS (Hybrid, grid feedback capable)

2. 5kwh 24v battery.

3. 1.2kwp of PV

4. AC connections: 

Main DB MCB -> Sub panel/DB MCB -> Inverter AC IN.

Sub panel/DB:

Inverter AC OUT -> MCB -> RCD -> Dedicated critical loads circuit.

I also have a Normally Closed contactor/relay, that does the N-G bond on the output, when grid is down (or MCB to inverter is OFF), since my inverter doesn't do N-G bonding automatically when in Solar/Battery mode.

Everything works fine, no nuisance tripping in any of the inverter modes (SUB/SBU). But up untill now, I have been using it with grid feedback disabled.

After enabling grid feedback, I noticed this problem:

If I turn off the input MCB (going to AC IN of the inverter), simulating a power outage, it causes the output RCD to trip immediately! This only happens when actively backfeeding. If I disable grid feedback, I can cutoff power to the inverter and everything continues to work fine.

After it trips, I can re-enable the RCD and it doesn't trip. I have tested this multiple times. 

I thought it might have something to do with double N-G bond on the output, but it shouldn't happen: When grid power is down (or I turn off the MCB going to AC IN), the NC contactor is de-energized and closes, making the bond.

Meanwhile the inverter's internal AC IN relay should open (I hear the click), so I shouldn't have another N-G bond from utility. Even so, it doesn't explain why this happens only when backfeeding.

Maybe when grid feedback is enabled, it takes more time for the internal AC relay to open? Does this make sense?

I gave this N-G bonding solution a great deal of thought and research on the forums, before deciding to go this way.

I didn't like the NEUTRAL to NEUTRAL bond solution, as it is controversial.

Any ideas? 

Thanks.

Just wanted to ask where did you buy the inverter from? I don't see that model online(offtopic)

15 minutes ago, Stefan Cornelissen said:

Just wanted to ask where did you buy the inverter from? I don't see that model online(offtopic)

https://www.mustenergy.co.za/product/ph1800-plus-series-2-5-5kw/ gives the information. 

May be just add ph1800 plus or pro to your search. Many sellers.

Edited May 13, 20242 yr by Scorp007

9 minutes ago, Scorp007 said:

https://www.mustenergy.co.za/product/ph1800-plus-series-2-5-5kw/ gives the information. 

May be just add ph1800 plus or pro to your search. Many sellers.

 

Mind dropping me the link? I only see the non grid feed ones listed

Mind dropping me the link? I only see the non grid feed ones listed

Edited May 13, 20242 yr by Stefan Cornelissen

10 minutes ago, Stefan Cornelissen said:

Mind dropping me the link? I only see the non grid feed ones listed

https://solarwarehousesa.com/products/3kva-mppt-pure-sine-wave-load-shedding-combo-3000-watt-2-56-kwh-ritar-lithium-battery?variant=47856312385857&currency=ZAR&utm_medium=product_sync&utm_source=google&utm_content=sag_organic&utm_campaign=sag_organic&gad_source=1&gclid=CjwKCAjw9IayBhBJEiwAVuc3foPOroGpwOJjOAVZQzykoEy04QdeOx5mgr9sB-xPSksuJFH4uSpQJhoCyu4QAvD_BwE

48V version. Not sure about the feed back. 

https://www.geewiz.co.za/stand-alone-inverters/194254-must-pv1800-pro-series-high-frequency-off-grid-solar-inverter-5200va-5200w-52kw-48v-450v-80a-mppt.html

Also have a look at the Must SA products link a posted earlier. 

Edited May 13, 20242 yr by Scorp007

2 minutes ago, Scorp007 said:

https://solarwarehousesa.com/products/3kva-mppt-pure-sine-wave-load-shedding-combo-3000-watt-2-56-kwh-ritar-lithium-battery?variant=47856312385857&currency=ZAR&utm_medium=product_sync&utm_source=google&utm_content=sag_organic&utm_campaign=sag_organic&gad_source=1&gclid=CjwKCAjw9IayBhBJEiwAVuc3foPOroGpwOJjOAVZQzykoEy04QdeOx5mgr9sB-xPSksuJFH4uSpQJhoCyu4QAvD_BwE

first one is the PV1800 VPM

48V version. Not sure about the feed back. 

https://www.geewiz.co.za/stand-alone-inverters/194254-must-pv1800-pro-series-high-frequency-off-grid-solar-inverter-5200va-5200w-52kw-48v-450v-80a-mppt.html

 

PV1800 Pro

 

Non of them does grid feedback only the plus version

first one is the PV1800 VPM

second one

PV1800 Pro

 

Non of them does grid feedback only the plus version

Edited May 13, 20242 yr by Stefan Cornelissen

1 hour ago, Scorp007 said:

This is how I understand what is mentioned in your post

I meant that during backfeeding, I switch off the AC IN. Of course backfeeding stops the moment I do this, because of inability to sense the grid. But it also trips the RCD on the output.

48 minutes ago, Stefan Cornelissen said:

Just wanted to ask where did you buy the inverter from? I don't see that model online(offtopic)

I bought it off Aliexpress about a year and a half ago. It took me a while to actually hook it up. Used it untill now as grid backup without PV. Now finally got to connecting PV.

This model is the PH-18 Plus, not very common. It's not available on Aliexpress anymore, but apparently still available on Alibaba.

Pro models usually have high MPPT voltage. PV-18 is totally off grid, not true hybrid as the PH-18 Plus. Mine has 145v MPPT max voltage.

 

 

Edited May 13, 20242 yr by meetyg

2 minutes ago, meetyg said:

I bought it off Aliexpress about a year and a half ago. It took me a while to actually hook it up. Used it untill now as grid backup without PV. Now finally got to connecting PV.

This model is the PH-18 Plus, not very common. It's not available on Aliexpress anymore, but apparently still available on Alibaba.

Pro models usually have high MPPT voltage. PV-18 is totally off grid, not true hybrid as the PH-18 Plus. Mine has 145v MPPT max voltage.

 

 

How much was the total cost to get it in SA?

1 hour ago, Scorp007 said:

You mention when you are feeding back when you switch the AC in off. Now the inverter should not be able to feed back as feeding back has to stop when there is no grid and it should actually feed back via the AC in which is connected to the grid side. 

It seems like the inverter is using PV to feed back via the back up side. 

The PV should only feed the back up load and battery charge when the grid is off via the AC in that you "trip" 

This is how I understand what is mentioned in your post. 

@meetyg what @Scorp007is saying makes sense your inverter will sense no uitility input if you switch the breaker off feeding your inverter input mains as soon as that happens either by a tripped input breaker or load shedding. What should happen when no utility input is sensed it opens the inverter ddu( bouble disconnect unit,) that isolates itself from the utilty and no feedback to the grid can happen for safety reasons the inverter then functions in isolated island mode only supplying power to the connected loads. Why the earth leakage is tripping on your output side.is a bit.of a head scratcher This is exactly why Coct test these inverter to ensure it isolates completely from the grid via your sdu(single disconnect unit)or ddu(double disconnect unit only then are they placed on the approve coct list.

 

 

 

 

 

 

1 hour ago, Stefan Cornelissen said:

How much was the total cost to get it in SA?

Sorry, I'm not from SA, rather from the Middle-East. But I found this forum very informative and it's members very educated and helpful.

So I hope it's OK that I'm participating here.

As opposed to other forums, where the majority of it's members are from the U.S.A, I find the information here more relevant to our local utility power system, including single-phase 220v systems. Also regulations are similar.

1 hour ago, TaliaB said:

Why the earth leakage is tripping on your output side.is a bit.of a head scratcher

I'm not convinced that it's backfeeding via loads/backup side. This would be a huge design flaw (not that those don't exist...).

After writing about my issue, I realized that I might be correct about the double N-G bond: If we assume that when backfeeding, it takes just a few milliseconds (or 10s of ms) to disconnect the internal AC IN relay, I could be in a situation with a double N-G bond on the output.

Maybe I was just lucky that when not backfeeding, the internal relay trigger is faster and so it doesn't happen in normal conditions.

I guess I need to figure out a way to delay the external N-G bond relay when grid is out. Maybe some type of AC capacitor?

I think I will test this first, to see if timing is really the issue. I will need to manually de-energize the N-G bond relay say a second or two after pulling the grid connection.

If I'm correct, I thought of getting a time-delay relay, that will delay for a few seconds. But I will need to rewire it, because they need a separate trigger signal from power. 

Edited May 13, 20242 yr by meetyg

So just an update, I think I solved the problem.

As I suspected my N-G bond contactor de-energized (creating the N-G bond on the output) before inverter released its grid relay. Just enough to create two N-G bonds and therefore the output RCD tripped.

So I got a "true" delay off timer. They call these "true" because it has some sort of internal capacitor to hold a small current for the delay time, when power is lost. 

Other off-delay timers exist, but they need an external source of power for internal timer to work.

The off-delay timer holds the N-G bond contactor powered (open) for a few seconds, giving the inverter time to release it's grid relay.

Now the RCD doesn't trip anymore when I simulate a grid-down situation by turning off the grid breaker.

Here is the model I got (JSZ3F with DIN-rail backplate):

https://www.aliexpress.com/item/32969613011.html

The only downside is that it's a bit bulky for the DB, had to cut out some of the plastic for it to fit.

 

Thanks for all of your inputs.

I hope this helps others too.