Surge Protection

[jason121]

Do any of you installers use surge protection on the DC side on commercial properties regardless if the building has LPS.

 

[SibertSolar]

Jason - I would hazard a guess that the general consensus here is a "no" unless Type2 devices are already incorporated in the inverter(s)

There are no regulatory requirements yet for DC surge and/or lightning protection - just best-practice advice if other factors are involved (LPS where separation distance is not achieved, or Type1 devices on the AC circuit, for example)

 

[jerriais]

Been at the Segen training today maybe? Lots of good advice there about that and many other aspects of commercial jobs.

 

[Gavin A]

Been at the Segen training today maybe? Lots of good advice there about that and many other aspects of commercial jobs.

are you not sharing the advice then?

 

[a1colly]

Over 50KWP systems then yes I have been fitting surge protection on bother the AC and the DC side of the installation. The main reason is for insurance purposes

 

[jerriais]

are you not sharing the advice then?

My advice - call the man at Dehn! Happy to share the slides when I get them from Segen. Or they might be on the website.

 

[Gavin A]

My advice - call the man at Dehn! Happy to share the slides when I get them from Segen. Or they might be on the website.

ah... I'd never take a manufacturers advice at face value, they are after all in the business of selling as many units of their product as they can, whether necessary or not.

eta obviously it's worth taking into account what the manufacturers are saying as well

 

[jason121]

The man from Dehn recommends that all commercial PV should a Type 1 device on every string. I worked it out very expensive to do

 

[jerriais]

True to an extent about not taking their advice at face value, but then by the same logic a customer could say the same about our advice (and often unfortunately that is what happens) even though I often find myself giving customers advice that actually results in us selling less panels, e.g. if another installer has recommended panels in an area of such severe shading that their return on investment would be negatively affected.
I guess I hope for an ideal world where at least some manufacturers and installers are giving advice that actually puts their customers best interests at heart! At least most on this forum including Gavin are certainly believers in this.

 

[Worcester]

Of course he does!!! (See GavinA's comments) there's a difference between recommended and required

 

[jerriais]

Of course he does!!! (See GavinA's comments) there's a difference between recommended and required

Certainly, and to his credit the Dehn guy was upfront about pointing this out.

 

[Gavin A]

my thinking on this really is that the inverters we install already have surge protection on the DC side in the form of varisters. Now fair enough they're use once only items, and would need replacing following a dangerous voltage spike, but realsitically how often is this going to be an issue, and if it is, then how much will it cost to buy and fit a replacement varistor vs the costs of fitting surge protection on every string on every commercial installation?

If they were that necessary then the inverter manufacturers would all fit them as standard in order to reduce warranty replacement costs.

So yes in areas of high risk of lightening strikes it's probably a sensible idea, but I'm really not convinced of the cost effectiveness of speccing it on most other situations.

 

[Worcester]

And the UK is a very low risk area - just do the risk analysis calcs for frequency of lightning strikes - parts of Germany, France, Spain, Austria, Italy are different.

 

[SibertSolar]

Guys, the risk analysis doesn't fully come into it on the AC circuit anymore, not under most of the listed scenarios in BS 7671 anyway. Since the Amendment back in June 2011, decisions on AC surge/lightning protection are still based on the risk assessment, yes, and are still returning a negative decision value based on the frequency of thunderstorms and lightning strikes (AQ Method and Ceraunic levels, other factors including overhead lines, buried cables etc come into play) BUT the risk assessment is over-arched if the installation meets any of 3 key criteria:

i) Consequences related to human life, eg: safety services, medical equipment in hospital
ii) Consequences related to public services, eg: loss of public services, IT centres, museums
iii) Consequences to commercial or industry activity, eg: hotels, banks, industries, commercial markets, farms

If any of the above applies then AC protection against overvoltage shall be provided (caveat - with consideration take to what's already fitted on the incomer, if anything, see below)

However, that all said, there is no such regulatory reference in relation to the DC circuits. For protection against surges induced on the DC circuit, relatively inexpensive Type2 devices (á la PowerOne Trio and SMA Tripower) are the solution - varistors are Type3 devices. However, the costs ramp up if the system is a ground-mount (under certain circumstances), or a roof-mount where a) an LPS is present, and b) the array frame cannot be separated far enough from the LPS to avoid "flashover". Whether to fit Type1 DC devices (budget for around £500 per string protected - whether individual, or post-combined strings) is simply a matter of subjective consideration based on the potential costs to replace damaged cabling and inverters in the event of a direct strike.

One other thing to mention about AC-side SPDs, if there is an LPS on the roof, then the odds are that there is a Type1 device on the main incomer. If this is the case, and you decide that the installation meets one of those 3 criteria above (or a risk assessment relating to overhead lines etc is undertaken and resulted in a need for SPD) then you would have to replicate the Type1 device at your circuit's grid connection point - fitting a Type2 device in this instance is about as useful as a chocolate fireguard....

Back to the OP's query - not many installations, even ground-mount, are being fitted with Type1 DC combined surge/lightning protection in my opinion. Not in the UK at any rate.

One last thing (sorry) to mention about SPDs - if you have RCDs in the circuit too then consideration should be given to their position in relation to the SPDs and whether a high surge-current rating (kA) version of the RCD is required. Manufacturer's default kA rating vary but the high surge-current rated units would typically be 3kA minimum

 

[Gavin A]

Guys, the risk analysis doesn't fully come into it on the AC circuit anymore, not under most of the listed scenarios in BS 7671 anyway. Since the Amendment back in June 2011, decisions on AC surge/lightning protection are still based on the risk assessment, yes, and are still returning a negative decision value based on the frequency of thunderstorms and lightning strikes (AQ Method and Ceraunic levels, other factors including overhead lines, buried cables etc come into play) BUT the risk assessment is over-arched if the installation meets any of 3 key criteria:

i) Consequences related to human life, eg: safety services, medical equipment in hospital
ii) Consequences related to public services, eg: loss of public services, IT centres, museums
iii) Consequences to commercial or industry activity, eg: hotels, banks, industries, commercial markets, farms

If any of the above applies then AC protection against overvoltage shall be provided (caveat - with consideration take to what's already fitted on the incomer, if anything, see below)

I'd disagree.

What would the consequences be in any of those situations if a lightning induced voltage surge were to knock out an inverter in any of those situations?

the consequences would be zilch other than they'd have to pay a little more for their electricity until the inverter would need repairing / replacing as far as I can see, therefore it's got no relevance to the situation.

All we're doing here is determining whether or not to protect the inverter from the direct impact of voltage surges or not, that's it, it will never have any serious knock on impacts that effect human life, public services or industrial activity.