With the change in technology from transformer type DC/AC inverters to transformerless types a problem has arisen with Sunpower rear wired solar panels.
We have used Sunpower pv modules for many years without degrading (abt 12). In order to prevent a build-up of static electricity on the surface of the panels we have connected a hard earth to the positive DC side of the inverter. This has worked well.
However with the introduction of transformerless inverters that means the AC system will be grounded directly to earth - ie a direct short. After consultation with many colleagues I propose to use a 'resistive' earth.
By measuring the current on my 30kW system I find that the voltage is in the order of 17 - 30 volts at a current flow rate of 0.001 milli amp.
It is proposed to place an in-line resistor of 0.5 Watt and 10 Mega OHM resistor with a 1 amp fuse for circuit protection. The connection will be made between the DC solar positive input and the MENS ground earth.
In that way the 'drain' will continue to function but with sufficient resistance to prevent detectable ground leakage from either the grid or the inverter.
I am soliciting comments before trialing the proposal in the field (and possibly frying an inverter).
Thank you for your kind consideration.
Les, if your inverters are truly transformerless then your resistor won't do anything because the inverter input - the PV - will be directly connected to the AC output. and allow any leakage current through won't it? But I think even 'transformerless' inverters contain high frequency transformers to adapt the PV voltage to the mains. So I don't see why you can't go on doing what you have been doing in the past. Which inverter do you plan to use? Is it really transformerless? Best regards, Tony
The inverter was a Fronius IG 60 (transformer type). It is planned to use a SMA SB 5.0 AV40. data sheet available at . It is assumed by us that there is a direct unbroken connection between the AC active and the DC positive. Therefore if the DC positive is hard wired (non-resistive) to earth then the AC active is essentially shorted directly to earth.
Is that a wrong assumption? Will the DC positive drain continue to function as it has in the past? If I understand you correctly do you mean that the 'transformerless' inverter will simply allow a drain of any leakage current without any further modification?
Best regards back,
Lex
Lex, I cannot believe that an internationally approved inverter can have an electrical path between the DC and AC circuits. This would risk having the entire PV array developing high AC voltages. It would fail isolation compliance testing immediately. I looked at the Sunny Boy information and the certification document certainly lists transformers made by a Vietnamese complany, and optocouplers made in Germany. I am very confident that you can continue using these inverters exactly as you have before with your earthing system. But to finally remove all doubt, just ask Sunny Boy if there is a DC path between input and output. If they say 'yes', choose another device IMMEDIATELY. Best wishes, Tony
I agree. It seems illogical that the loss of an neutral would have the effect of electrifying an entire pv frame (as the earthing system would become the neutral).
For that reason we have put the 'DC earth' on its own system and away from the AC MENS.
I will check with SMA. To keep the question simple 'Is there a direct, unbroken path between the DC and AC circuits of SMA inverters?'.
Tony Maine
I have sought clarification from SMA as to DC/AC path. However the statement from GSES clearly states that there can be a current flow between the two systems. .
Thanks for that.
Lex
Les, I read your document and it does allow under certain conditions a non-isolated inverter though not with an earthed array. In my personal opinion the guys who wrote this should be stood against a wall. This is a crazy situation, you could blow the inverter to pieces and put the full mains voltage onto the array if the inverter fused into a short circuit. I would very very strongly recommend NEVER to use an unisolated inverter and your problems will disappear. Designers of inverters should pay attention to all possible fault scenarios. I dont believe that in this case they have.
For clarification - Industry is changing out the rear wired panels for front wired, which do not require the positive earth. In your opinion, is that simply delaying the inevitable?
Ted Spooner are you able to help with this? I am very uncomfortable with the situation. It is a matter that I think should be resolved. On the one hand we could be throwing away very serviceable panels (which renewable industry might not mind), on the other hand we could be heading into a serious problem (which the fossil fuel industry might not mind). Be Aware the 'Pink Batt Debacle'.
kind regards
Lex
PS this is a real problem I have a client with an existing pv array (transformered inverter since 2008) who now requires a new inverter. Transformerless inverters are all that are available for systems over 3kW.
Les, I think there's an issue with definitions here. Make sure that the 'transformerless' inverter means a mains-frequency transformer. All good inverters should have a high frequency transformer to provide galvanic isolation between the AC and the PV. I am confident that Sunny Boys do, but you should get the manufacturer to state this. Never use an unisolated inverter on the mains. Reserve them for low power off mains use only.
SMA Australia advise a clear NO any modification will void the warranty. They have declined to confirm a solution with Germany. They advise that 'reverse flow' is prevented by a system of diodes.
I am posing the question to Fronius.
The solution of replacing the Sunpower panels with front wired ones, which do not require the 'functional earth' appears to be industry's solution. It flies in the face of responsible environmental policies if we are throwing out fully servicable panels if for no reason than no one can be bothered addressing the issue.
Accountants refer to that as 'technical obsolencence'. The DC 3 remains a fully servicable aircraft yet no one wants to fly them anymore. - such a waste.
That still does not address your concern of AC flow to the frames.
I will keep you posted.
Regards
Lex
Just tell SMA that until they confirm they have full galvanic isolation from DC to AC you won't be buying or recommending any of their products. They will change their tune quite quickly. On the other hand, you might not bother and just use someone else's product anyway. A manufacturer that refuses to confirm their designs conform to good engineering practice isn't worth even considering as a supplier.
I agree. I have already gone to Fronius. They are far more co-operative. It is a bit sad as we have had a very good run with SMA in the past. Their 'old' inverters are very robust and well suited to the harsh conditions of the North West.
The CEC goes some way to addressing the issue of earthing with 'transformerless' inverters and do raise the matter of 'resistive earths'. However I cannot find any detail on it and am now more than suspicious of a resolution.
It appears the answer to the problem from both SMA and Sunpower is that the are happier to sell new panels but appear disinterested in resolving technical issues. I guess that is the world where economic factors outweigh social and environmental ones. I could write a book on that - oh that's right - I have '[T]Axing Greenhouse Gases, due for release next year.
Thank you so much for your help. We have spent far more time on this than we should have.
Mind you I still have the problem with my up grade.
Kind regards
Lex
Breaking News. Fronius have suggested the simplest way is to place a galvanic DC - DC transformer between the array and the inverter. They are checking that and will get back.
A DC to DC transformer is just a high frequency transformer/rectifier combination which could and should be far more easily integrated into the inverter. Where oh where is the understanding of solar PV systems these days? What do inverter manufacturers really think they are doing? I agree with you Les, this is all taking too much time, and it should not even be an issue in 2018!!!
- Badges
- Science topic