High-wattage solar modules increase risk of thermal runaway

[sunshine_eggo]

Top level:

High-wattage solar modules increase risk of thermal runaway

Longi Solar outlines its high-temperature mitigation logic in designing the lower current, high-wattage Hi-MO5 solar panel series.

www.pv-magazine.com

Detail:

LONGi- Analysis of Thermal Runaway in Large Current Junction Boxes

Advances in photovoltaic technology are closely related to improvements in module efficiency.

www.longi.com

Given that it's from a company offering a product, a certain level of skepticism is needed, but it passes a logic sniff test - high current generates heat. it presumes that high current is bad, and it should be addressed in design. It implies higher current panels may carry a higher risk of thermal runaway, but that presumes that those panels do not address this increased potential in its design.

My conclusion: Meh. Given the hardware commonly used, panels that put out more than 10A can complicate installations especially as users choose higher Voc MPPT with lower PV input current limits (18A seems common). This puts a mental ~10A limit on panels I'll consider.

 

[BradCagle]

Yeah, so basically as modules become more efficient the need for higher capacity junction boxes, and more robust diodes. I mean that's logical.

I've always worried about the bypass diodes. Consider a string of 6 panels in series, one or more of these panels gets hard shaded, the bypass diodes kick in. What happens when one of these diodes fail catastrophically (pops), and a high voltage arc forms across the terminals.

Do you think the cells in that run will pull the current back enough to extinguish the arc?

 

[foggysail]

BradCagle-- I am not sure where you are going to get a high voltage across a bypass diode, high enough to cause an arc. This does not mean you are wrong, I just cannot conceptualize such a condition. Can you elaborate please. My recollection of the voltage drop across an arc is in the 50V range so yes, I agree that an arc can occur in a dc solar system but not across a diode because there is not enough voltage there.

There are examples of arcing in string inverter systems but those arcs require enough voltage to initiate an arc at the time when current is flowing. Sure, if there is any inductance in series with a dc wire break an arc can easily be established even if the voltage seems small because at the time of the wire break in an inductive circuit the voltage will soar based on the equation V = L*(di/dt) but again, after the initial incidence there must be enough voltage for the arc to continue and I remember that to be in the 50v range.

 

[BradCagle]

BradCagle-- I am not sure where you are going to get a high voltage across a bypass diode, high enough to cause an arc. This does not mean you are wrong, I just cannot conceptualize such a condition. Can you elaborate please. My recollection of the voltage drop across an arc is in the 50V range so yes, I agree that an arc can occur in a dc solar system but not across a diode because there is not enough voltage there.

There are examples of arcing in string inverter systems but those arcs require enough voltage to initiate an arc at the time when current is flowing. Sure, if there is any inductance in series with a dc wire break an arc can easily be established even if the voltage seems small because at the time of the wire break in an inductive circuit the voltage will soar based on the equation V = L*(di/dt) but again, after the initial incidence there must be enough voltage for the arc to continue and I remember that to be in the 50v range.

Honestly I don't know if this could happen, just a thought. But consider if the diode is the conductor (or path of least resistance) ATM because the cells are hard shaded, the conductor burst open, like a breaker would. Is it possible for an arc to form? In this scenario what is the difference if you cut the diode in half, or just the wire, or a switch, or disconnect a MC4?

The only thing I can think of is the diode is in parallel with a string of cells, and perhaps this will keep the arc from forming, or if it does it will extinguish rapidly because the cell string becomes the new path of least resistance. Now that I'm sounding this out, I "think" this is the case.

 

[foggysail]

Brad, there has to be enough voltage across the break for an arc to continue. If you could put a voltage meter across the arc itself you would read a voltage in the 50V range. So if not enough volts here, you will not have a continuous arc. But CAUTION here! You certainly can get an arc pulse. Remember I mentioned inductance my equation it does not take much inductance.

 

[BradCagle]

Brad, there has to be enough voltage across the break for an arc to continue. If you could put a voltage meter across the arc itself you would read a voltage in the 50V range. So if not enough volts here, you will not have a continuous arc. But CAUTION here! You certainly can get an arc pulse. Remember I mentioned inductance my equation it does not take much inductance.

I agree. However what if the string of cells is damaged, lets just assume they have a totally broken circuit. Now the voltage across the diode is still not that much provided the diode is still in one piece. Now what happens if the diode ruptures to where it's split in two halves? If you measure the two terminals of the diode now you will see full voltage the series array string.