WARNING! Major Issue With 18Kpv, Exporting to Grid for hours when Not Programmed To!

[fnnwizard]

Well, my call with the tech specialist goes nowhere. I need a senior Engineer to speak with. And please listen to that phone call. I've been more help to SS than they have to me... Really. Cmon SS.

Anyway, I took a closer look at the log data and something is definitely not right.

In the BAR graphing menu it, shows the total energy drawn from the battery, but in the spreadsheet log and line graph, it does not show battery discharging.
In AC Charging mode, 1st priority is energy from grid AND no discharging from battery, period.

The Spreadsheet log is NOT completely correct.

It's obvious that if there was exporting to grid and with my sub panel constant load running; the energy has to come from the battery at 4AM.

My BMS confirms a draw of 600-1100W during the exporting period, the Bar charts show Battery discharging, yet nowhere in the spreadsheet logs or the line graph does it show.

So my question is where are the data being polled from. Again, it's very obvious the spreadsheet data is wrong in a lot of the columns.
Look closely at the spreadsheet info. I've hi-lited the important columns for the coders to really take a look at.

I can for sure say I still have a constant inverter load of 200-400W. This is because I have network gear and surveillance systems running 24/7 and there is no reports of power issues.

My only objective is to make this 18Kpv work like it is described in the white sheet or manual ( hopefully both).

So man, let me speak to someone that can make the correct changes in the code or tell us, it can not work the way We all thought it would.

 

[sunsurfer]

Well, my call with the tech specialist goes nowhere. I need a senior Engineer to speak with.

How's your mandarin?

 

[SignatureSolarJess]

Well, my call with the tech specialist goes nowhere. I need a senior Engineer to speak with. And please listen to that phone call. I've been more help to SS than they have to me... Really. Cmon SS.

Anyway, I took a closer look at the log data and something is definitely not right.

In the BAR graphing menu it, shows the total energy drawn from the battery, but in the spreadsheet log and line graph, it does not show battery discharging.
In AC Charging mode, 1st priority is energy from grid AND no discharging from battery, period.

The Spreadsheet log is NOT completely correct.

It's obvious that if there was exporting to grid and with my sub panel constant load running; the energy has to come from the battery at 4AM.

My BMS confirms a draw of 600-1100W during the exporting period, the Bar charts show Battery discharging, yet nowhere in the spreadsheet logs or the line graph does it show.

So my question is where are the data being polled from. Again, it's very obvious the spreadsheet data is wrong in a lot of the columns.
Look closely at the spreadsheet info. I've hi-lited the important columns for the coders to really take a look at.

I can for sure say I still have a constant inverter load of 200-400W. This is because I have network gear and surveillance systems running 24/7 and there is no reports of power issues.

My only objective is to make this 18Kpv work like it is described in the white sheet or manual ( hopefully both).

So man, let me speak to someone that can make the correct changes in the code or tell us, it can not work the way We all thought it would.

We are the distributor.. would you like me to get you connected with EG4 to see if they can assist any further?

 

[EG4TechSolutionsTeam]

Last Thursday, my 18kPV decided on its own to export from ESS to grid in the middle of the night.

I have a temporary test setup to diagnose some odd behavior related to PV Priority Charging and got bit by a major issue with the EG4 18Kpv exporting power to the grid for a few hours! Yes I am upset!!!

I've spent all weekend trying to figure out how can this be and unless I can get into the firmware, I'm lost. Called in to SS multiple times before this with PV Priority Charging issues - which does not work even remotely close to the way it is described in either the manual, or the EG4 White paper and was always told "we'll have higher up level support call you back"... I'm still waiting for ANY call back.

I don't know how many levels of support there is but I can't wait on this EXPORT ISSUE. So hopefully can get any/all expert feedback here.

I viewed this as so critical that I would have thought EG4/SS would have reached out to me by now, but even with my multiple calls today and having other tech and salespeople leave what I assumed were written notices, I'm left wondering what are they doing... arhggg.

Anyway, here are the setup and logs.

Main panel is outside, w/ a couple circuits w/ constant load 120-500 watts to mitigate export spikes.

50A breaker from main panel feeds L1 and L2 to "Grid" connection of 18kpv, breaker located at bottom rails. N bonded with G here. CT's on L1 and L2 here, arrows pointing to "Grid" connection of Inverter.

Sub panel fed by Load of inverter L1,L2 into 50A breaker. N and G separated @ Sub panel.

Sub panel has AC (17A@240V), EV (16A@240V), Microwave ( 20A@120V), Air Fryer (15A@240V), other misc circuits to test load up to ~12KW if all running simutaneously with 1 of those circuits always on, pulling around 400W.

So during the period in question, the inverter should be pulling close to 600W FROM the GRID ( from sub and main panels), not exporting to it.

It was pulling that at 10pm, 11pm on the 18th and into the 19th so was working correctly on the 18. Nothing was changed on the 18th to make it behave differently.

Application Settings are below and remains unchanged.

I had AC Charge Enabled, but AC Charge Power set to "zero" during the export.

What gives?

@Markus_EG4
@SignatureSolarJames
@Luxpower_Gilbert
@EG4TechSolutionsTeam

Edit to add wiring diagram

From viewing the data, it appears to only be showing export wattage during charging cycles. Could we try moving the CTs in-between the utility meter and panel to see if this corrects the readings?

 

[fnnwizard]

How's your mandarin?

Non Existent, but there are lots of workarounds for this. Point me to the right person.

We are the distributor.. would you like me to get you connected with EG4 to see if they can assist any further?

Of course. How about getting someone from there to call me. Like I said, it's in all of our best interests to fix this.

From viewing the data, it appears to only be showing export wattage during charging cycles. Could we try moving the CTs in-between the utility meter and panel to see if this corrects the readings?

View attachment 230910

Here's the thing you are not understanding... this was a random, completely unexpected event. If this was a CT issue, it would have done this a long time ago. From my testing, the CT reports the current flow to the inverter. The inverter does nothing with this data as far as controlling the DIRECTION.

You want to test this out, just enable reverse on the CT SETTINGS and see the load info show up as Export to grid in the spreadsheet column. Even though it will log this flow as amount exported, it is not a true export if you actually measure power back to grid.

I tried this last night for about 30 min. You try it and you will see. Too much info to be typing out here. Suffice to say the inverter does not stop that flow either.

The principle issue here is:
1. 18Kpv export when:
a. not programmed too
b. explicitly in a state to pull grid power
c. explicitly in a state to not discharge battery

The software itself is contradicting the relevant data.

2. Did not catch the fact that it was exporting.
a. CT were working correctly, showing true exporting and not just backwards install.
b. something tried to stop the exporting between 1am and 2 am, but proceeded for the next 4 hours to export.

The CT showed the correct export amount from Inverter. It works perfectly fine!

The main panel loads consumed some of that energy so true export to grid shows a bit less. PoCo data has it right.

I don't really care what I pull from the grid. I absolutely care about every little bit of energy I EXPORT from the inverter! That means best practice is to have CT right at GRID INPUT OF INVERTER.

A little export spike here and there is expected. I accounted for that. But for a straight 4 hours? C'mon..

The CT is a 'dumb" data collector. It will not differentiate between the wire I have it on now to the one right after the service meter other than it would show a constant import load of what remains in the main panel, thereby offering inaccurate export data if there were continuous export of , 300-400 watts from the inverter.

Do you see and understand this?

Edit. The CT might stop exporting if Export was enabled and the Export energy limit was reached. It may then stop, but I doubt it from all the other sub menu (branch) oddities I have found.

I make a lot of edits since I make a lot of typos. When dealing with writing, I try to be as accurate as possible grammar wise, since.. a simple omission of a comma can steer the meaning of a sentence.

 

[SignatureSolarJess]

Of course. How about getting someone from there to call me. Like I said, it's in all of our best interest to fix this.

Looks like they've joined the thread

 

[EG4TechSolutionsTeam]

Non Existent, but there are lots of workarounds for this. Point me to the right person.

Of course. How about getting someone from there to call me. Like I said, it's in all of our best interests to fix this.

Here's the thing you are not understanding... this was a random, completely unexpected event. If this was a CT issue, it would have done this a long time ago. From my testing, the CT reports the current flow to the inverter. The inverter does nothing with this data as far as controlling the DIRECTION.

You want to test this out, just enable reverse on the CT SETTINGS and see the load info show up as Export to grid in the spreadsheet column. Even though it will log this flow as amount exported, it is not a true export if you actually measure power back to grid.

I tried this last night for about 30 min. You try it and you will see. Too much info to be typing out here. Suffice to say the inverter does not stop that flow either.

The principle issue here is:
1. 18Kpv export when:
a. not programmed too
b. explicitly in a state to pull grid power
c. explicitly in a state to not discharge battery

The software itself is contradicting the relevant data.

2. Did not catch the fact that it was exporting.
a. CT were working correctly, showing true exporting and not just backwards install.
b. something tried to stop the exporting between 1am and 2 am, but proceeded for the next 4 hours to export.

The CT showed the correct export amount from Inverter. It works perfectly fine!

The main panel loads consumed some of that energy so true export to grid shows a bit less. PoCo data has it right.

I don't really care what I pull from the grid. I absolutely care about every little bit of energy I EXPORT from the inverter! That means best practice is to have CT right at GRID INPUT OF INVERTER.

A little export spike here and there is expected. I accounted for that. But for a straight 4 hours? C'mon..

The CT is a 'dumb" data collector. It will not differentiate between the wire I have it on now to the one right after the service meter other than it would show a constant import load of what remains in the main panel, thereby offering inaccurate export data if there where continuous export of , 300-400 watts.

Do you see and understand this?

Edit. The CT might stop exporting if Export was enabled and the Export energy limit was reached. It may then stop, but I doubt it from all the other sub menu (branch) oddities I have found.

I make a lot of edits since I make a lot of typos. When dealing with writing, I try to be as accurate as possible grammar wise, since.. a simple omission of a comma can steer the meaning of a sentence.

Do you mind DM’ing me the serial number for the inverter so that we can take a look?

 

[fnnwizard]

Do you mind DM’ing me the serial number for the inverter so that we can take a look?

Serial sent. I don't mind, but please don't change anything without letting me know first.

Regarding the CT info, look at the data when I reversed its setting at 11:43pm my time last night. If you just spend a few minutes really analyzing the data you can easily come to the conclusions I have.

And by the way, I am still waiting on feedback regarding the PV Priority Settings. This White paper is closest to describing how it works, but still far, far from reality. The only right description is that as soon as PV Priority is enabled, all loads are powered by grid. after that nothing works the way it should.... reality is ALL loads will be powered by Grid no matter what... even when batteries are 100% charged, pv is completely out of the equation.

But that's not as pressing as export issue.

 

[Hedges]

While I always like to blame firmware/software, I can also think of a hardware reason why it would randomly switch to exporting.

If one CT was backwards, or wires swapped, or leads connected wrong internally, then imbalance between L1/L2 phases would make it toggle between seeing import and seeing export.

50A breaker from main panel feeds L1 and L2 to "Grid" connection of 18kpv, breaker located at bottom rails. N bonded with G here. CT's on L1 and L2 here, arrows pointing to "Grid" connection of Inverter.

Checking the manual, I see that orientation is correct.

I was going to say if you got the wires backwards, that would be same as reversing CT. The one's I'm using with Wattnode have white and black wires. But 18kPV manual shows RJ jacks.

Make sure you plugged each jack into the right socket (or each CT is on the right wire.)

Is it possible to examine what color wire goes to each pin of CT? Could have been assembled wrong.

By running an experiment with no house loads except a space heater on L1, then space heater on L2, you could confirm both register as power draw.

If one of L1, L2 was reading backwards, then imbalanced loads could cause inverter to seek zero export with incorrect outcome.

(Just focusing on hardware, because that's so much easier to address than firmware/software.)

Check voltage L1-N, L2-N. If imbalanced, that could possibly imbalance power draw and what is seen as zero.

 

[Eddie_LuxPowerTek]

Markus@EG4 is on this now. It’s unusual case, let’s see what’s happening in the end. Thx for reaching out to Luxpower office today @fnnwizard

 

[Quattrohead]

While I always like to blame firmware/software, I can also think of a hardware reason why it would randomly switch to exporting.

If one CT was backwards, or wires swapped, or leads connected wrong internally, then imbalance between L1/L2 phases would make it toggle between seeing import and seeing export.





Checking the manual, I see that orientation is correct.

I was going to say if you got the wires backwards, that would be same as reversing CT. The one's I'm using with Wattnode have white and black wires. But 18kPV manual shows RJ jacks.

Make sure you plugged each jack into the right socket (or each CT is on the right wire.)

Is it possible to examine what color wire goes to each pin of CT? Could have been assembled wrong.

By running an experiment with no house loads except a space heater on L1, then space heater on L2, you could confirm both register as power draw.

If one of L1, L2 was reading backwards, then imbalanced loads could cause inverter to seek zero export with incorrect outcome.

(Just focusing on hardware, because that's so much easier to address than firmware/software.)

Check voltage L1-N, L2-N. If imbalanced, that could possibly imbalance power draw and what is seen as zero.

You have brought up an excellent point, we are talking about 4 little wires into an RJ45 plug and if one pair were reversed this could happen. Simple mistake to make during crimping of the RJ45. Or maybe one pair isn't crimped properly or open circuit for whatever reason and it's only actually able to read one CT.

 

[Markus_EG4]

Last Thursday, my 18kPV decided on its own to export from ESS to grid in the middle of the night.

I have a temporary test setup to diagnose some odd behavior related to PV Priority Charging and got bit by a major issue with the EG4 18Kpv exporting power to the grid for a few hours! Yes I am upset!!!

I've spent all weekend trying to figure out how can this be and unless I can get into the firmware, I'm lost. Called in to SS multiple times before this with PV Priority Charging issues - which does not work even remotely close to the way it is described in either the manual, or the EG4 White paper and was always told "we'll have higher up level support call you back"... I'm still waiting for ANY call back.

I don't know how many levels of support there is but I can't wait on this EXPORT ISSUE. So hopefully can get any/all expert feedback here.

I viewed this as so critical that I would have thought EG4/SS would have reached out to me by now, but even with my multiple calls today and having other tech and salespeople leave what I assumed were written notices, I'm left wondering what are they doing... arhggg.

Anyway, here are the setup and logs.

Main panel is outside, w/ a couple circuits w/ constant load 120-500 watts to mitigate export spikes.

50A breaker from main panel feeds L1 and L2 to "Grid" connection of 18kpv, breaker located at bottom rails. N bonded with G here. CT's on L1 and L2 here, arrows pointing to "Grid" connection of Inverter.

Sub panel fed by Load of inverter L1,L2 into 50A breaker. N and G separated @ Sub panel.

Sub panel has AC (17A@240V), EV (16A@240V), Microwave ( 20A@120V), Air Fryer (15A@240V), other misc circuits to test load up to ~12KW if all running simutaneously with 1 of those circuits always on, pulling around 400W.

So during the period in question, the inverter should be pulling close to 600W FROM the GRID ( from sub and main panels), not exporting to it.

It was pulling that at 10pm, 11pm on the 18th and into the 19th so was working correctly on the 18. Nothing was changed on the 18th to make it behave differently.

Application Settings are below and remains unchanged.

I had AC Charge Enabled, but AC Charge Power set to "zero" during the export.

What gives?

@Markus_EG4
@SignatureSolarJames
@Luxpower_Gilbert
@EG4TechSolutionsTeam

Edit to add wiring diagram

After checking on this myself and then double checking with the engineers there is absolutely no way you exported. No export at all. The inverter was in AC charge mode at that time, with AC charge power set to zero. I think there was a calculation error when the power was really small.

If you receive a confirmation from your utility company that you did and if they fine you we will take care of the charges 100%. But there is no way you sold back.

This would be the first case ever if you receive anything from your utility company we have had

 

[Solar Guppy]

The only way I see these external values happening is the CT's are reporting energy movement. I'm with @Hedges on this one, the CT's are either incorrectly installed, possibly defective or noise in the sense wires going to/from the CT and inverter.

Also, what is the purpose of not having the CT's on the feed to the main panel breaker? You can't even zero out the other home loads on the main panel the way its installed on the 50 amp feed in breaker to the inverter.

I'm not familiar with this inverter, but if you are only selling solar and can sell all, I would removed the CT's ( disconnect the RJ45 ) and see what if anything changes, could help pinpoint the cause if software or hardware.

 

[fnnwizard]

After checking on this myself and then double checking with the engineers there is absolutely no way you exported. No export at all. The inverter was in AC charge mode at that time, with AC charge power set to zero. I think there was a calculation error when the power was really small.

If you receive a confirmation from your utility company that you did and if they fine you we will take care of the charges 100%. But there is no way you sold back.

This would be the first case ever if you receive anything from your utility company we have had

Yep, thats what I had first thought too, see my post #30. I understand AC Charge priorities. I am 100% sure it was true export. Did you take a look at barcharts of grid export?

There is always a constant draw on my main panel so PoCo Data always shows a draw. But not so on the 19th at 5am and 6am.

Here are the Poco data from 16 to 22.

My 3rd party CT, which is on the grid wires verifies the export.

 

[Markus_EG4]

Yep, thats what I had first thought too, see my post #30. I understand AC Charge priorities. I am 100% sure it was true export. Did you take a look at barcharts of grid export?

There is always a constant draw on my main panel so PoCo Data always shows a draw. But not so on the 19th at 5am and 6am.

Here are the Poco data from 16 to 22.

My 3rd party CT, which is on the grid wires verifies the export.

Most of these seem to look like surges from a HVAC unit. I do see the data from the 19th from your external CTs going in and out as well. I really hope it didn’t sell back but if it did we will cover anything you need and provide what we can to your utility company.

I have never seen this before especially while AC charge is enabled. There’s got to be something else more info we can’t see yet.

 

[fnnwizard]

Most of these seem to look like surges from a HVAC unit. I do see the data from the 19th from your external CTs going in and out as well. I really hope it didn’t sell back but if it did we will cover anything you need and provide what we can to your utility company.

I have never seen this before especially while AC charge is enabled. There’s got to be something else more info we can’t see yet.

No, There is only 1 verifiable surge and that is on the 5th at 9pm. And was very lucky to have caught it since that log polls every 3 minutes.
And I know exactly why that surge showed up. It was due to battery being in standbye so wasn't quick enought to take the "charge"

For surges to happen, there has to be large loads in the 2 cell column to the right of pToGrid, immediately before the surge. So there may be lots of surges that are never caught due to the 3 minute polling intervals -= gives perhaos a false sense of security.

Also, the data on the 23 showing export is not True export. I did that to show what reversed CT looks like and can easily be verified with other data in the logs.

 

[Markus_EG4]

There’s always an answer. Just need to dig more. What is POCO I’m trying to look up your other device you mentioned picking up sell back as well?

 

[fnnwizard]

There’s always an answer. Just need to dig more. What is POCO I’m trying to look up your other device you mentioned picking up sell back as well?

poco is power company = Socal Edison

3rd party CT just some cheap similar to this from amazon

 

[Quattrohead]

3rd party CT just some cheap similar to this from amazon

Cheap and frequently returned item according to Amazon and the reviews are all over the place.
You need to use a revenue grade meter to substanciate your claim of random backfeed.

 

[fnnwizard]

Cheap and frequently returned item according to Amazon and the reviews are all over the place.
You need to use a revenue grade meter to substanciate your claim of random backfeed.

Really? That's all you have to contribute? Would you consider the power company's meter revenue grade? Or maybe you don't know how to read a simple graph.

For someone who just less than 3 years ago asks about energy conversion units... you sure are puffing your chest a weeee bit in this thread here yea?

That Tuya device is great! I'd suggest doing a bit more research on it before writing something you know nothing about.

You know, with someone like you, all the data in the world could hit you squarely in the face and you would continue to deny it.

But honestly, so far, your remarks have been the most worthless in this thread IMNSHO. I'd normally say get out of here... but if you really want to learn something, shutup and stick around.

 

[Brucey]

Cheap and frequently returned item according to Amazon and the reviews are all over the place.
You need to use a revenue grade meter to substanciate your claim of random backfeed.

I've heard that eg4/sig solar requires jumping through a bunch of hoops to get warranty satisfaction but this seems ridiculous.

 

[wpns]

I've heard that eg4/sig solar requires jumping through a bunch of hoops to get warranty satisfaction but this seems ridiculous.

Well, Quattrohead isn't making policy for EG4, but there's no way you need revenue grade equipment to know which way power is flowing.

It sounds like EG4/SS are finally responding with the appropriate people, let's not jog their elbows till they and fnnwizard figure out what's going on. It's sure to be something no-one thought of, from an invisible wiring error to a well-hidden firmware issue (remember blowing up stuff when using the crossover CAT5 cable?), but I have faith in both sides that they'll be able to figure it out and report back.

 

[TomC4306]

Really? That's all you have to contribute? Would you consider the power company's meter revenue grade? Or maybe you don't know how to read a simple graph.

For someone who just less than 3 years ago asks about energy conversion units... you sure are puffing your chest a weeee bit in this thread here yea?

That Tuya device is great! I'd suggest doing a bit more research on it before writing something you know nothing about.

You know, with someone like you, all the data in the world could hit you squarely in the face and you would continue to deny it.

But honestly, so far, your remarks have been the most worthless in this thread IMNSHO. I'd normally say get out of here... but if you really want to learn something, shutup and stick around.

Wow, lashing out
Sad ...
Now I don't give a f*** what the outcome is.
Here's a tissue.

 

[Quattrohead]

Well now Mr excitable, you come on here all guns blazing with a massive attention-grabbing heading and then later on admit you're using a piece of shit to measure stuff. And then lashing out at me just blows away any reputation that you may or may not have, so let's all stick around and get this figured out.
By the way it wasn't me that pissed in your Wheaties this morning.

 

[1201]

Well now Mr excitable, you come on here all guns blazing with a massive attention-grabbing heading and then later on admit you're using a piece of shit to measure stuff. And then lashing out at me just blows away any reputation that you may or may not have, so let's all stick around and get this figured out.
By the way it wasn't me that pissed in your Wheaties this morning.

Do you have personal experience with the energy meter he is using? Or just going by Amazon reviews?