diy solar

diy solar

Designing panel/MPPT controllers around extreme weather?

KevinC_63559

New Member
Joined
Jan 26, 2024
Messages
125
Location
NE Missouri, USA
If I'm following design rules, I believe I need to factor in record weather to come up with an appropriate correction factor correct?

e.g. Although our average minimum temperature is around -9C, its not uncommon to have a week or so at -20C. Looking at record temps, we have gotten down as low as -35C. This year, for example, was REALLY cold (at least for us) with about 10 days around -32C.

Found this table in some old Midnight documentation I had stored:
1709843151303.png

Feels like I should use 1.25 to be safe, given the unknown variation of upcoming climate change. Does that make sense?

Of course, I'll do the math based on whatever panels I eventually select in conjunction with whatever MPPT charge controller I select maximum's VOC.

Its just that planning for the worst known conditions, when all one has is historical data, feels dicey. It would be painful to blow up an MPPT (and potentially burn up the building its in), just because we had a random really cold sunny morning.

Does anyone other than Midnight offer anything similar to their HyperVOC (which will shutdown their MPPT units if the maximum voltage is exceed, with a non-operating range boosted by that plus the battery voltage (e.g. Using one of their 250V chargers with a 48V battery allows for up to 298V before the unit is damaged)? Seems like that would be quite handy for extreme weather a few days a year.

For example, if I used a Victron 450V/100 Smart solar with (8) 50VoC panels in series I'd be fine most of the time, (down to around -5C) but would fry the system on that cold day (400X1.25=500 >> 450). To be somewhat safe I'd have to limit the system to (7) panels (427.5Voc).

Am I doing that math right?
 
Looks like same data, but maybe more current:


You should look at the actual solar panels as they have a Voc temperature coefficient. IIRC, the NEC tables correspond to about -0.4%/°C. MANY panels are higher than this (less negative) providing more margin.

You have the concept/math correct.
 
My slightly better than NEC numbers panels saved my ass. Pre-forum and n00b time for me saw 3S 72 cell panels on a Victron 150/100 because I didn't know better. Fortunately, that winter was mild, and I recognized my error before damaged occurred. Upgraded to 250/100, and all has been well.
 
Yes, use record coldest temperature.
As Sunshine said, use actual temperature coefficient for your specific panels.

I use 1.16x as a safe rule of thumb (I'm not in Sibera, haven't see numbers where 1.25x would be needed), then sharpen the pencil if I want to cut it closer than that. That was based on -15C vs. 25C, and 0.4%/degree temperature coefficient.

A table from Midnight ... Midnight Classic is one that can survive over-voltage, just doesn't operate. I think all they do is not start switching, so the transistor sees Voc - Vbat instead of Voc. I think that might not work so well if battery was disconnected.
 
I've seen -22 to -28 twice in the last 20 years and I'm in Central Ohio! It doesn't stay long but it's always clear and sunny when it does.
 
A table from Midnight ... Midnight Classic is one that can survive over-voltage, just doesn't operate. I think all they do is not start switching, so the transistor sees Voc - Vbat instead of Voc. I think that might not work so well if battery was disconnected.

I concur, their website claims its Voc + battery voltage. No battery hookup === no battery voltage.
 
Back
Top