Basic fusing question (maybe)

[FilterGuy]

And, what kind of shut-off to use, with the highest voltage rating.

The disconnect should disconnect both the positive and negative, so it needs to be two pole.

The IMO line of DC disconnect switches is the Go-To for many folks, but they tend to be pricy.

Some people will use a dual pole breaker as a disconnect switch because they tend to be lower cost than a properly rated DC disconnect.
If you go that route, the breaker rating should be high enough that it will never trip. (Something larger than 1.56 x Array Isc)
I don't like using a breaker as a switch unless it has a very high cycle life rating. Furthermore, by the time you buy a box to mount the breaker in, you may not be saving all that much.

 

[FilterGuy]

What are you trying to protect against ?
1/ Current back flow.
2/ Over current.

Use the appropriate cure for the problem you hope to solve.

Take your pick... If a parallel panel gets a short, there needs to be something to prevent overcurrent due to backflow from the other parallel panels. Either way, there is no rule on how a fuse or a diode should be used. They both have a certain set of characteristics and the designer must choose the best device for the application. Back in the analog days, lightbulbs were often used as resisters in industrial applications. Not because light bulbs make light but because their characteristics also happen to make them the best high-power resistors for the application so they were the appropriate cure for the problem.

I also noticed the discussion is no longer about whether there can be problems with a short (I guess we agree there can be). We are now discussing how to deal with the problem. I don't know why diodes are not accepted by the code for preventing the problem but I suspect it has to do with their reliability characteristics as a safety device. If they are not properly cooled, they will tend to fail. Even if they are properly cooled, they are more likely to fail than a fuse. Even if they fail 'open', it still means someone has to service a high-voltage line that can't easily be turned off. Meanwhile, a fuse will only fail in response to something else failing. The bottom line is that the NEC requires fuses or breakers (not diodes) and pretty much all of the solar industry uses fuses or breakers. (There was a time when I thought I knew better than the NEC, but after eating a few slices of humble pie I got over that.)

Note: There is nothing in the code that prevents the use of blocking diodes. However, the code does require OCPDs for more than two strings of panels.

Note 2: Blocking diodes are a relic from when people put '12V Panels' directly onto the terminals of a 12V battery. (Yes.... they did that.) When using a modern charge controller, blocking diodes are rather useless. (This is another little factoid that I had to eat humble pie on) For the longest time, I thought blocking diodes were needed if you had strings facing different directions to capture morning and afternoon sun. It turns out they are not needed because even a shaded panel can create Voc if there is no current. Consequently, a shaded string will create a high enough voltage to prevent backflow from the other strings. (this is a different scenario than a shorted panel)

 

[Warpspeed]

You are right about failure modes.
Diodes do tent to fail shorted, and fuses fail open circuit.

Diodes are probably one of the most reliable components, and oversizing them for reliability is never really an issue.
However, if code says something must be done to pass, that is the end of it.

 

[mtts]

Note: This conversation has strayed off-topic. I believe the OP was asking about surge protection, not fusing. I also believe he said there is only one string.

The problem is my question was one from my own broad ignorance. It is not easily answered.

There is general PV to equipment protection (1), at the first level. Thank you for everyone's extremely thorough answers.

For (2), those of you unfamiliar, 'Professor' Ben Davidson has written several books. He's professor-level but very relatable, not wrapping terms in unnecessary levels of difficulty to protect a job, career field, etc. {He could be a prof at a uni, but has chosen instead to get his message out directly. One of his books is used at unis.}

I have watched coronal mass ejections/flares at the X100+ level leave the sun - not in earth's direction. That is a 'kill shot' that would fry the grid, and do damage to off-grid systems. Although I don't know how to protect from it.

They are not yet at the scale of the solar micronova predicted between 2035-2045, which is why Bezos and Zuckerburg are building the giant underground centers, and why the Air Force has been involved in building similar structures since the 1940's. One of those X100, roll of the dice, may hit earth between now and then, anyway.


This micronova is much more severe, at a once-in-12,000 year event level. It is the blow-off from an accretion layer, that is built-up from sol passing through the center of the galactic wave-patterned sheet. Out here, the wave has longer, more open oscillations. Every 3,000 years there is some level of interaction. Every 6,000 years a stronger event, and every 12,000 (12,068) a much stronger event: the so-called Black Sun. Where there is enough microparticulates in the wave, to envelope the sun, which it eventually blows-off, leaving the 'pale rider', a yellow-green colored sun. The 'hell follows with it' part is the shell buildup hitting the earth, with the micronova blast.

Ref:
Thomas Chan, Adam and Eve Story (uncensored scan) https://libgen.rs/book/index.php?md5=F3FC7FA0973FC52E973047B181CF5263
Ebook version https://libgen.rs/book/index.php?md5=85DEF30907FEA6E6174EE28E87C210F8
World In Peril, by Ken White, grandson of Major Maynard E White:

Full text of "World In Peril The Origin"

 

[mtts]

If you follow Ben as I do your worried more about EMP or coronal mass I'm guessing. Do a search on EMP. Best thing to do is have a back up
in a faraday cage.

Yes. I cleaned up my grammar above, so it's more readable. I hear you, and wonder if putting extra panels inside a connex, with some sort of disconnect protection rated to handle that X100 blast to the panels (not the larger micronova: nothing will survive that, save being underground in a cave), would be enough?


For those of you that are like, ? 'oh great, another skitzo thread': I've put sources above. The solar micronova outburst skews carbon dating tremendously, making it inaccurate to get a real rate, on anything that went through the last cycle. It could also be the reason why we find evidence of past advanced cultures, but no steel tools; all of that, at least on the surface, would have been vaporized in the induction wave phenomena associated with the micronova.

 

[timselectric]

So this is one of those really bad things that we should all be freaking out about. But there's no proof of it. Because when it happens, it conveniently destroys all proof that it ever happened.
Got it!!

 

[mtts]

So this is one of those really bad things that we should all be freaking out about. But there's no proof of it. Because when it happens, it conveniently destroys all proof that it ever happened.
Got it!!

How to tell me you've never reviewed the evidence, without telling me.

Next

 

[Warpspeed]

Either a Carrington level event, or a full on weapons grade EMP shot, are both so severe there is very little any of us can do about damage prevention.

Its kind of like asking what sort of roof tiles should I fit to my house to resist the direct hit by a bomb ?
You need to start thinking about yards of reinforced concrete, not roof tiles.

Even a direct hit by lightning will pretty much destroy what it hits every time.
A nearby strike is something you might be able to offer some protection for.
The best strategy is just make sure your solar panels are not the highest exposed object in the general area.
If they are, nothing is going to save them.
Some kind of grounded mast structure to divert the lightning away from your panels might be a worthwhile first step.

An EMP or Solar flare is a very different thing, and realistically there is a point beyond which nothing is going to prevent damage or destruction.