I have a question that I am not in a position to answer right now, so perhaps someone may be able to help.
My envisaged system is in a locale that to me is borderline when it comes to overvolting a 150V MPPT. Whether this is a hard threshold or not, I treat it as one.
So I have up to now, only used two 72 cell panels in series, not 3. I may standardize on 60 cell panels in future, but then it would be 3 panels in series not 4.
I want to cheat the big difference in voltage between the Vmp and the Voc, if I can.
And this is where I wonder if I can, because I know several of you have systems on the coast and can trust that you will never have frost, and can take the risk of 3 x 72 cell panels in series.
My completely baseless untested hypothesis is that I will only get freezing conditions as well as sufficient light in the early morning.
Can I get away with that extra panel in series, on North and West strings? - Things should’ve warmed up by the time they start to deliver.
(If I stick to 2 panels in series on East strings).
I have no historical data, to either support or disprove this. So I hope that a few board members can chip in with the sort of maximum voltage they see early on cold mornings.
P.S.
( I don’t even know if systems record the PV voltage before the MPPT kicks in, or if the MPPT kicks in very early and avoids these higher open-circuit voltages altogether).
All my Victron installations consist of either 3 series 72 cell panels on a 150 volt mppt ot 4-5 72 cell panels on a 250mppt. I have experienced that the Mppt connects to the PV long before the panels even comes close to the VOC drawing it down and keeping it way clear of VOC values.
I am currently driving to Gauteng for work. Will get some readings for you when I reach my destination later tonight. Are you interested in coastal readings only or will readings from all over work.
I’m going to say no. Strictly speaking, just because they point away from the sun, doesn’t mean they don’t still get light. As I understand it, a PV module’s current production is affected by light levels more than the voltage it makes, so even indirect light at a cold temperature will be sufficient for it to make a too high a voltage.
I cannot prove this right now, this is just a gut feeling based on the mechanics of things. For any current to flow, you need enough energy to kick an electron out of a silicon atom’s orbit, and any amount of energy that is enough to kick electrons out (even if it’s less kicking than would happen in full sunlight) is likely going to need a particular potential difference, that is to say, I don’t think you can bargain on a lower voltage just because it’s not in direct sun.
So far I still think the neatest “cheat” I’ve seen is the Trina smart modules. They have a built-in clamp in the module, which simply clamps the voltage down to a lower level, thereby reducing the difference between Voc and Vmp.
I agree with what you say, I read the VI curves exactly like that.
Your logic is absolutely theoretically faultless. This is the practice I have adopted. It hasn’t failed me and if someone did it on the coast it wouldn’t fail them either.
The thing is over-engineering costs money. Those copper losses are every day. I don’t like that.
I have been told since way before the internet that it is theoretically impossible for bumblebees to fly, but they do.
I don’t mind bumbling along if it means I can fly.
I am trying to guard against a 30-year low temperature which could be 50 years from now or it could be next Tuesday. Chances are that historical low temperature, happens during the night, the historical records don’t say.
This is where science and art meet.
Where does shelling out moola, and practical risk meet?
Testing the theory has the potential to become extremely expensive, but I understand the MPPT starts doing its thing 5V above the battery voltage.
If that’s true, then freak low temperatures and MPPT failures become a double failure and an acceptable perfect storm risk I am willing to accept.
I am happy to operate in theoretical grey areas, if practice beats theory. I’ll push it to the limits, but I also accept that limits are limits.
If peoples’ data say it’s not the way to go I’ll be a disciple, but if practical data shows otherwise I am happy to ride the whirlwind.
On a normal day, I understand the MPPTs are drawing down the voltage. However, what would happen when 1. your AC breakers are open (maybe you are working on the power/tripped because short circuit, and 2. the batteries are full, maybe even also disconnected.
In that scenario the MPPTs have no ability to start drawing current to lower the voltage or do I misunderstand?
In that case, I think the normal working conditions is one consideration, but a second would be (if you are going to spec your system close to what you think it can handle) to make sure you always open the fuse from the panels to the MPPT when you suspect a no load condition might exist.
This might be difficult to control, especially if your inverter shorts on the output. Maybe the batteries stays connected but they might also be full. Or perhaps their BMS disconnected as well.
Not sure how likely sure a scenario is, but I guess it might be part of the “overengineering” side of your question.
@jykenmynie
Again, the theory supports what you say.
What I am trying to determine is practically what actually happens?
Is there a practical lag of the voltage rise time between East and North or West strings that can be exploited?
The theory says that a panel achieves full voltage with very little light, I accept that.
Then again, I recall MPPTs switching off a fair bit before it was actually dark in the afternoon when I was drawing power.
I would expect that an East string MPPT string would have switched off earlier than a West string for example.
I don’t know if that actually happened. I didn’t pay enough attention at the time. I was not questioning this aspect as that layout design was finalized.
These are the practical considerations that raise my query.
This is a future system, so the design is back on the drawing board.
60cell panels are not always that easy (or cheap) to get anymore, although they are available. I search for my PV panels so that their max voltage were within the limits of my MPPT.
Not all 400W panels get to that power using the same voltage. I found that Canadian Solar have a higher max V, while Art Solar had a lower max V with more current to for their 400W panels.
So here is what I see on my Art Solar 425W panels.
Their spec:
And over the last 2 winters these were my absolute max for 2S2P strings on a 100/20 MPPT
I am with you there. I too am entirely willing to take ignore that once-in-30-years low-temperature event and make the strings slightly longer. But I also know the MPPTs I use have at least some protection built-in. They will at least attempt to protect themselves above 145V. So as you say, it becomes a double failure thing.
Please tell me more about this, in my next build I will be using Victron MPPTs. I have already acquired them. @Louisvdw, Thank you very much for your historical data. It would seem that you could run a 3S string and only hit 145.45 volts.
From memory, I only theoretically run into issues at -2 degrees C in the existing system. That temperature is just about possible, but that is about as cold as I’d really expect it to get.
I used Artsolar panels,I am guessing you are about the same. -2 deg C puts me just on the high side of 150V, (in theory). I’ll take the risk if, in practice, it actually has to get 2 deg C colder than that.
I don’t have full details on this. As I understand it, the MPPT will stop charging and clamp the voltage above 145V (for the 150V models). So essentially it just “shorts” the panels and allows the full current to run in a circle without doing any work, until the voltage drops.
So in theory, if the voltage shoots up from below (as it would when the sun comes up), the MPPT would be able to catch it in that window and prevent damage. Even if the voltage isn’t linear to the light level, it’s also not a square wave that goes on like a switch in the morning
But if you suddenly turned on a breaker and exposed the MPPT to an immediate square-wave type jump over 150V, it might not be able to do anything about it.
I’m about 5km or so from the ocean. The coldest I saw was an actual zero, but there was no frozen water or damage to the direct solar water heating setup, so it must have been a mild zero, so to speak.
I calculated at the time that I’ll be okay above -7°C. The last time the Western Cape was below that was 50 years ago, or something like that…
Going forward with 5 in series at my peak after 12pm at -4deg with cloud effect … these Canadians I have are quite potent when they peak at like ±5.5kw from a 4.2kw array on nice cool days with cloud effect.
So yea, each to their own, I don’t trust the “weather” anymore.
It is changing “fast” internationally.
If it actually does this, that would be sufficient enough to take the chance in my mind.
Coupled with the fact that the cold temperature would be a morning event and I am very unlikely to have no loads and full batteries after the night.
