Cable sizing and breakers for solar

[hellosailor]

If you exceed the ratings (voltage or amperage) of a breaker, the contacts may arc over and flash weld themselves closed when you wanted them to open and break the circuit.

This is one reason why batteries are usually fused, not "breakered". A standard 12VDC breaker can carry about 3000 amps before the contacts weld shut. Seems like way more than you would ever need, right? Except, one Group 24-27-31 SLI battery can put out over 3600 amps into a crowbar short (i.e. a wrench shorting it out at or near the battery) so in fact you need a more expensive 5000A rated breaker. Which makes fuses very much more affordable and popular.

The general rule with electrical ratings is that gadgets physically break down, catch fire, slag down, etc. when the ratings are exceeded. If they are rated conservatively and well made, not so critical. (Perhaps.) But generally? You may want to stay 25-50% UNDER the ratings, to give yourself some added safety margin.

[Dsanduril]

As Noelex says, breakers can get complicated (very) when you move beyond 12 or 24V DC. This is because the voltage is constant and can easily develop an arc (with AC the voltage goes to zero 50 or 60 times a second and the arc dies). At higher voltages the arc is longer, which means the gap created by the contacts in the circuit breaker has to get larger. That's the technical side.

So, look for a breaker that is truly rated above your voltage. NoArk makes their Ex9BD series that is rated to 250V DC in single pole and 500V in two-pole. Anything rated above about 80V would be fine (Eaton, for instance, makes a bunch of these).

But then you also have to look at the "tripping curve", which is where exactly the breaker will open (and you'd think that would be simple - 10A means it opens at 10.01A).

A 'C' curve, which is very typical needs at least 6 times its rated current to open in magnetic ('fast') mode, your other two strings of panels can never provide that much, so it will never open in magnetic mode.

A 'C' curve will trip at it's rated current (+ up to 20% tolerance) in 10000 seconds (2 hours 45 minutes) on the thermal ('slow') curve. The guaranteed thermal trip point of a 'C' curve 10A breaker would be 12A, and only after almost three hours. Your other two panel strings could - in theory - contribute 12.5A into a fault on the third string and that may - or may not - trip a 10A breaker after a very long time.

Thus, breakers on small arrays with few parallel strings/panels usually do not provide much protection. They're simply not designed for that service. If you have 4 (barely) or more strings in parallel then they start to make sense, but at less than that they are mostly just a disconnecting means should you want to disconnect the panels for some reason.

[Edit] - If you truly want overcurrent protection (OCP) on the panel side of things then the only thing that can provide it is something like an 8A fast acting fuse. That should open just over 8A and would provide protection. Of course, now you have a non-resetable device with contact surfaces that seem to corrode easily. [/Edit]

[Dsanduril]

Originally Posted by hellosailor

.... Except, one Group 24-27-31 SLI battery can put out over 3600 amps into a crowbar short (i.e. a wrench shorting it out at or near the battery) so in fact you need a more expensive 5000A rated breaker. Which makes fuses very much more affordable and popular.....

I neglected this part of the breaker question as I was describing things, for both breakers and fuses the other part of the selection process is the 'interrupting capacity' or 'short circuit breaking capacity'. Both fuses and breakers have interrupting ratings beyond which they can fuse/melt/catch fire as Hello described. You have to know the largest fault current that your system can provide to make this selection. And, yes, you should have a bit of overcapacity here as knowing the available fault current can be a bit difficult.

Lithium batteries are terrible in this regard, even small ones can provide immense fault currents and need very robust OCP. Solar panels, on the far side of the controller (the controller, plus the fuse between the controller and the batteries should block faults from the battery to the panels) have very low available fault currents (they simply cannot provide more than their Isc) and thus don't need a large interrupting rating.

[john61ct]

Originally Posted by Dsanduril

Lithium batteries are terrible in this regard, even small ones can provide immense fault currents and need very robust OCP.

Only if you are designing to **allow** such high currents.

If you calculate the highest current xA you **want** to flow through a circuit, then why pay more for fatter wire?

But even if you do, the fuse can be sized for 25% above x anyway.

[Dsanduril]

Originally Posted by john61ct

Only if you are designing to **allow** such high currents.

If you calculate the highest current xA you **want** to flow through a circuit, then why pay more for fatter wire?

But even if you do, the fuse can be sized for 25% above x anyway.

There are two ratings on a fuse (and circuit breaker, etc.) - the current at which it is designed to blow (trip for CB) and the maximum current it can handle before it melts/fuses/blows up (yes, fuses can literally blow up spewing molten metal and plastic or even glass).

For a 12V AGM battery with a typical 0.5mOhm internal resistance the available fault current is around 4000-5000A. That means your 50A or 100A or 200A battery fuse has to be able to withstand a 5000A current and do its job without being utterly destroyed (yes, the fuse element is destroyed, but the case contains the blown element). A fuse with a 10kA interrupting rating will handle this easily.

A 12V LiFePO4 battery may have an internal resistance of 0.1mOhm. That means that it can deliver ~5x the current of the AGM, or 20kA to 25kA available fault current. The same 10kA IC fuse that is used for AGM batteries could turn into a melted nightmare if used for LiFePO4 - you need a 30 to 50kA IC for use with lithium.

(All of the above being simplified and not including connection resistance, cable resistance, etc. that must all be considered in proper design).

Doesn't matter what your design is in terms of load amps, it's all about how much fault current the batteries can put into the fault, if the batteries can put in more than the IC rating then the chances are that either you will have catastrophic failure of the OCPD or you will have a sustained arc across the OCPD. Neither of these is fun.

[McG]

Wow, a very BIG "Thank YOU" to everyone that replied.
I am now so confused I am going to outsource the job .... just jokin'

Thanks for the detailed info. I am not an ocean sailor, I am a land sailor in an RV but this is the best place to get info. I love this forum and the people.

Cheers to all!

[noelex 77]

Originally Posted by john61ct

Only if you are designing to **allow** such high currents.

If you calculate the highest current xA you **want** to flow through a circuit, then why pay more for fatter wire?

But even if you do, the fuse can be sized for 25% above x anyway.

Dsanduril is referring to the IC or (AIC) or interrupt capacity of the fuse or breaker, this is quite different to current rating of the breaker. The correct IC rating is dependent on the batteries capacity to deliver high currents.

The IC becomes especially critical in lithium systems.

There are guidelines for reasonable IC ratings in marine systems based on lead acid battery size. I am not sure if any have yet been published for lithium systems, but they will need to be considerably higher.

It is something that needs consideration especially when upgrading from lead acid systems. I suspect insurance companies are nervous about Lithium batteries because in some installations details like this are neglected.

[doublebubble]

Hi, I'm about to add another 400w of flexible panels to my existing 450w, my existing panels are all hard framed with built in diodes, but I'm not sure if flexible panels have blocking diodes or not, I've been looking at a few panels of interest but there is no mention of diodes, can anyone suggest a good flexible panel around the 200w mark, I will be using a Victron mppt controller.

Regards Rod.

[noelex 77]

Originally Posted by doublebubble

Hi, I'm about to add another 400w of flexible panels to my existing 450w, my existing panels are all hard framed with built in diodes, but I'm not sure if flexible panels have blocking diodes or not, I've been looking at a few panels of interest but there is no mention of diodes, can anyone suggest a good flexible panel around the 200w mark, I will be using a Victron mppt controller.

Regards Rod.

Very few panels have blocking diodes installed. All (or nearly all) panels will have internal bypass diodes installed as these are necessary to protect the cells from overheating.

Blocking diodes are easily installed, but for most installations they are counterproductive. If you want to connect panels in series adittional bypass panels are probably helpful.

[doublebubble]

Noelex77 thanks, I get a little confused about which is the best way to configure panels, parallel or series, which is the best harvesting method using mppt controllers, or is there a best method, at the moment I have 1 x 250w panel with is own controller and 2 x 100w panels in parallel on there own controller, any suggestions is highly appreciated.

Regards Rod.

[noelex 77]

It would be great to see more practical tests,but generally the view is that from best to worse, assuming the panels are the same and 36 cells or more is:

Single panel per controller
Parallel connection
Series connection

The Victon controllers have an unusually high start up voltage. So with some low voltage panels series connection may be better in some circumstances. I suspect these circumstances are rare, but there are very few practical reports to help clarify this.

[doublebubble]

Ok thanks for that, its clearer now as I'll leave what I have the way it is and just add two more flexible panels but unsure as to which ones, brand to purchase, keep googling.

Regards Rod

[Group9]

Originally Posted by john61ct

Certainly not true for Victrons. Their 75/15 is designed so that 40+V is much more efficient than "12V" lower voltage panels.

Right up to 65V, then their 150 versions fine with 140V.

Something about giving the MPPT algorithm more headroom to find optimal power point, that's how they get 15+% better output than PWM.

And this is all well before, nothing to do with the buck conversion.

Thats my understanding. MPPT controllers can make use of the extra voltage, while PWM controllers just shunt it off as heat.

[noelex 77]

Originally Posted by Group9

Thats my understanding. MPPT controllers can make use of the extra voltage, while PWM controllers just shunt it off as heat.

Yes that is true (appart from the bit about heat) , but “12v” panels are already producing around 17+v in average conditions. An MPPT controller can convert 17v to the battery voltage and therefore produce more current than a non MPPT controller.

Series conection inceases the voltage differnce further, but this hurts rather than improves the efficiency converting the solar panel voltage to battery voltage. In simple terms for most circuits the higher the voltage difference between input and output the lower the efficiency. However, the real loses of series conection are in partial shaded conditions.

There is a very slight gain for series under very low light conditions, but the power produced by panels, in low light, that are below battery voltage (which is where the gain occurs) is tiny so this does not generally make up for the overall shade losses.