3rd inverter in less than a year, what could be wrong?

[mokoono]

Hi,

I had a new 14 panel (approx 5.6KW) system installed last December with a Growatt SPH5000 inverter, no batteries by an "award winning" installer.

Two inverters have failed, both due to power cuts and I am now organising a third replacement with the company. I suspect there might be an underlying fault in the install or with this model of inverter that means a 4th replacement isn't too far away, but I have no idea what that would be? Nothing else in the house blows up, and neighbours solar inverters have survived the same power cuts for years.

A bit of background first on this troublesome install:

• Installed and "commissioned" in the dark in December, next day when sun hit the panels tripped out with PV High warning
• Turns out they should have split into two strings as the single string voltage exceeded the max DC voltage.
• After many delays, cancellations they sent a "roofer" in February to restring the panels who had to borrow my crimping tools (no, not proper MC4 ones), he also told me there was no electricity in the DC/Solar cables, it was just data!
• Had to assist the confused roofer with my own multi-meter and neither string worked. More delays, the electrician arrives and discovered the polarity had been reversed, swapped - working system at last!
• One month later in March, power cut, inverter stops working, Growatt remote in and confirm fault. Protracted return via Growatt before a replacement install in April. Assumed the old inverter had been compromised by the previous miswiring in some way.
• Working great for 4 months, exporting to Octopus, using home assistant to monitor etc... Until today - Short power cut, inverter completely dies, even the display is dead, no remote diagnosis on that!

Two inverters killed by short power cuts, probably not a coincidence. I am not an electrician (my dad was, retired) but I am OK with basic electronics and IT, but not enough to diagnose this.

Company is coming out next Thurs, hopefully to fully replace but that's not entirely certain at this point, they like to drag things out. Should I be asking them to check some other parts of the install before sacrificing another inverter?

Any advice and thoughts welcome.

Thanks, Matt.

 

[binky]

bit of an odd one, any idea what cuased the power cuts? Could be voltage spike from the grid, might be wrth getting them to put a surge suppressor on the incoming supply / board.

 

[mokoono]

The power cuts are fairly regular here and the surrounding villages, once or twice a month. I think I read there is some expensive investment needed that they're avoiding. UK Power Networks reported a "high voltage overhead electricity line fault", affected about 1000 homes in an 8 mile radius.

I think you're right, it was probably a voltage spike. If the installer won't cover a suppressor, sounds like it might be worth me asking my local electrician to look at that.

Just strange that no other equipment was affected and neighbours solar installs are fine, I would expect an inverter to be able to handle a power cut. Maybe this model is particularly sensitive to voltage spikes.

Thanks for the suggestion.

 

[binky]

You can get some very strange behaviour on electrical systems, dependent on where you are on the grid. A lighting strike on the outskirts of Plymouth a few years ago wiped out 2 inverters on completely different estates, with a large wooded valley in between them. It made little sense and neighbours inverters were fine.

 

[ProDave]

Turn your ac isolator on and off. That will simulate a power cut. If the inverter dies when you do that then it is a lousy design of inverter.

If multiple operations of the ac isolator don't kill the inverter then it is likely it was more than a simple "power cut" that killed it.

 

[mokoono]

Good idea for a simple test, as you said, it would a pretty poor design if it couldn't handle being turned on and off like that.

It certainly had no problem being turned being turned on and off normally in the past, obviously never stress tested it by repeatedly switching the isolator.

My suspicion is some sort of voltage spike caused by the power cut as Binky said, rather than just the power being cut off.

Not sure if I want to try that test when I get the new one, maybe I'll get the installer to try it instead

Thanks for the suggestion

 

[binky]

Good idea for a simple test, as you said, it would a pretty poor design if it couldn't handle being turned on and off like that.

It certainly had no problem being turned being turned on and off normally in the past, obviously never stress tested it by repeatedly switching the isolator.

My suspicion is some sort of voltage spike caused by the power cut as Binky said, rather than just the power being cut off.

Not sure if I want to try that test when I get the new one, maybe I'll get the installer to try it instead

Thanks for the suggestion

they are desgined to handle power loss, hence the UPS function I believe your inverter has.

 

[UNG]

The power cuts are fairly regular here and the surrounding villages, once or twice a month. I think I read there is some expensive investment needed that they're avoiding. UK Power Networks reported a "high voltage overhead electricity line fault", affected about 1000 homes in an 8 mile radius.

I think you're right, it was probably a voltage spike. If the installer won't cover a suppressor, sounds like it might be worth me asking my local electrician to look at that.

Just strange that no other equipment was affected and neighbours solar installs are fine, I would expect an inverter to be able to handle a power cut. Maybe this model is particularly sensitive to voltage spikes.

Thanks for the suggestion.

It sounds like you are in a rural area so I'm wondering if your power supply delivered overhead, had a fault similar to this about 30 years ago that would cause random equipment to fail at the customers premises and there was no apparent or traceable reason, the fault was eventually found after the customer borrowed a mains analyser from the company he worked for this found that there was an occassional 1100v spike lasting about 5 - 10ms that coinsided with an appliance failing, Norweb subsequently did some checks and found no issue but their logging kit was not sampling quick enough to see the fault which was eventually traced to a problem with a local transformer

 

[mokoono]

Our power supply isn't delivered overhead, despite being rural its a large village and I live on an estate. However, I wouldn't be surprised to find there was occasional spike or surge. 1100v sounds crazy high and I would expect other equipment to fail, like my servers, routers, tvs etc. That said, most of them do have a basic surge protector on them.

I think I've had such a poor experience with this solar installer I was expecting to find it was something they'd done making it more susceptible. It's sounding more like these strange supply spikes and surges move in mysterious ways, and I could just be unlucky. If only I had access to one of those mains analysers

I am going to contact them today and see if they are open to installing some surge protection and to confirm they ARE actually replacing the inverter next week.

System had been saving us over £100/month recently, last time it failed I was without a working system for over 2 months!

Thanks for the advice.

 

[binky]

You can rent an analyser or talk the DNO and try to persuade them to install one. It's worth checking the inverter itself to see what it thinks the grid voltages are running at. Because the grid is underinvested in, quite often the DNO is raising voltages to cope with extra demand, which is grounds for them to install an analyser. It won't be the first time a solar install has found issues with the grid, although I suspect grid spikes are the issue, caused by whatever is cutting the grid supply.