GFI fault showing on Power One Inverter

[silverpulser]

I have a nominal 4Kw system comprised of 16 REY solar panels arranged on my roof with 11 facing south east and 5 facing south west.

it is now 11.15 am and the Gfi (Ground fault indicator) LED is still illuminated. The total generation (FIT) meter is showing a red light which means that no power is being fed into the grid (the same as at night time). I am now going up into the loft to attempt to restart the system and photograph the Aurora Power One (model 3.6) Inverter screen whilst it is still displaying the fault message.

OK, I have now completed my observations and they are as follows:

When I first looked at the fault message on the screen it read “Measuring riso (see attached image)”. This message stayed on screen with the salt timer icon revolving continuously.

I then turned off the two individual panel strings and waited for the Inverter to turn off. I then turned on the 1st string (facing south east) and it eventually showed a message which read “Riso low” which settled down to a reading of 0.60 M (see attached image). The fault message remained on the screen and pressing the “Esc” button only illuminated the screen and did nothing else.

I then turned off the 1st string and waited for the inverter to turn off. I then turned on the second string (facing south west) which started up normally with a Riso measurement of 20 M (normal). Whilst this was still operating I turned on the first string and that too started up normally with both panel strings showing a good output of 1750w and 700w respectively.

I then turned off both strings and waited for the inverter to turn off and restarted both strings simultaneously with the result that the original Gfi fault showed and the “Riso low” message displayed. I turned off the two strings and once again started them up consecutively by first turning on the 2nd string (facing south west) until it showed me the normal Riso of 20 M and outputting around 700w then turning on the 1st string (facing south east) which also now started up normally with an output of 1750w. This is how I have now left the system.

The temperature reading on the inverter was 12.5C when first observed and 14C when I last observed it.

What is the most likely cause of this GI fault?

Shaun

 

[SolarCity]

It is, unfortunately, not an uncommon problem. The most likely cause of it is water getting into one of your MC4 connectors on the rear of the panels. It is something which your installer will need to put right.

The good news is that it is probable that the fault has occurred on the end of the string which means that correcting it isn't as complicated as it might be.

Winter storms trigger low insulation resistance (RISO) issues | Power-One

 

[silverpulser]

It is, unfortunately, not an uncommon problem. The most likely cause of it is water getting into one of your MC4 connectors on the rear of the panels. It is something which your installer will need to put right.

The good news is that it is probable that the fault has occurred on the end of the string which means that correcting it isn't as complicated as it might be.

Winter storms trigger low insulation resistance (RISO) issues | Power-One

OK, I get that but why does it start up OK if I start the good string first? Incidentally in that case the Riso tends to show much less than the 20 Mohms, usually somewhere between 3 and 7 but still generates without a warning light because It is not less than 1 Mohm.

Shaun

 

[BruceB]

Because it only tests the insulation resistance before it starts up. So if you start up successfully using the good string you can then add in the leaky string and it will keep going.

 

[silverpulser]

Because it only tests the insulation resistance before it starts up. So if you start up successfully using the good string you can then add in the leaky string and it will keep going.

OK guys, thanks for your input. The bad news is that after talking to the installer I have to wait for his son (a qualified electrician but not in solar installations) to return from attending a wedding abroad before getting it tested. Also, although the fault displayed for over a week, it has not done so for the last 36 hours.

Shaun

 

[silverpulser]

OK guys, thanks for your input. The bad news is that after talking to the installer I have to wait for his son (a qualified electrician but not in solar installations) to return from attending a wedding abroad before getting it tested. Also, although the fault displayed for over a week, it has not done so for the last 36 hours.

Shaun

Installer still hasn't been to investigate or even returned message left on the answerphone.

The Riso today is showing as 1.8 Mohm Will this effect output? It is midday and sunny at the moment and the output from my 16 panel (4Kw) rig is 2.8 Kw. Is that about right for this time of year. I have 11 panels facing south east and 5 panels facing south west.

Shaun

 

[liquidity]

Installer still hasn't been to investigate or even returned message left on the answerphone.

The Riso today is showing as 1.8 Mohm Will this effect output? It is midday and sunny at the moment and the output from my 16 panel (4Kw) rig is 2.8 Kw. Is that about right for this time of year. I have 11 panels facing south east and 5 panels facing south west.

Shaun

I would guess the installer is not exactly looking forward to installing scaffolding, getting back up on the roof, removing the panels and checking the connections one by one...

I believe it's a binary situation - either the inverter works 100% or it detects one or the other (or both) of the isolation resistance and current leakage values being out of range and refuses to start.

 

[silverpulser]

Does the low resistance effect output?

Is 2.8 Kw about right for my rig at this time of year

Shaun

 

[Gary K]

My 3.84, all SSW is peaking at about 3.1Kw so you are about right IMO.

 

[silverpulser]

At last the installer is sending an electrician around this afternoon. So I went into the loft to measure the Riso on the faulty string and now, because of the recent sunny warm weather, the Riso is measuring 20 Mohms. Obviously the moisture that was causing the problem has now dried out so where does that leave me? The two year warranty (which cost me an extra £300 to extend it from one year) expires at the end of this month (March) and I am feeling somewhat exasperated that this was not dealt with when it arose and when the fault would have been more easily identified. If this fault arises again in the next prolonged wet spell where do I stand as the warranty will by then have expired.

Shan

 

[Ian-LS]

You paid for a two year warranty extension? with the installer?

They are supposed to give you at least a 2 year warranty to start with and shouldn't charge you for an extended warranty to take it from one year to two years. This is detailed with the REAL scheme I believe.

 

[silverpulser]

Solar electrician has now been and gone. As expected there was nothing to find after all the recent sunny weather. He did note that the connectors weren't MC4 but rather Phoenix connectors and they were located inside the roof space. Not sure what that all means but in any event he has asked me to call him if/when there is any change in the Riso measurement.

Shaun

 

[silverpulser]

Solar electrician has now been and gone. As expected there was nothing to find after all the recent sunny weather. He did note that the connectors weren't MC4 but rather Phoenix connectors and they were located inside the roof space. Not sure what that all means but in any event he has asked me to call him if/when there is any change in the Riso measurement.

Shaun

Update:

Scaffolding erected and electrician inspected the suspect string. One connector was wet and changed. Remainder tested OK.

For the week following the RISO has measured between a low of 8.1 Mohms and high of 15.8 Mhms at startup. On each occasion it returned to 20 Mohms when restarted later in the morning.

Electrician rang Power-One who say this is normal where there is morning dew.

What do you guys think?

 

[Gavin A]

the limit is usually 1 Mohms for BS7671 (wiring regs), so it's a bit low, but passes. It's normal to have a lower RISO figure when there is dew or in wet weather due to reduced resistance / increased capacitance from the panels (IIRC).

 

[edexlab]

Heard of this problem with the connectors but have yet to see it for myself, most PV connectors are IP 67 so a bit of morning dew or even heavy rain shouldn't be a problem surely?
unless the ends haven't been tightened up properly, the tools for this are rubbish!