I have a MultiPlus-II 48/5000/70-50 connected to a hydroelectric turbine, with the AC IN limited to 16 A (approx. 3,500 W), in order to supply my household electrical loads.
The energy storage system consists of two Pylontech US5000 batteries.
The “overload” alarms are triggered at around 8,000 W (once or twice a month).
Based on my understanding of how the MultiPlus works:
Up to 50 A entering through AC IN can be delivered through AC OUT.
The inverter can “provide” a peak of 9,000 W and 4,000 W continuously, in addition to what comes in through AC IN.
Is it possible to calculate the “Overload” threshold (based on our installation)?
As Thierry said, it already depends on the temperature of your inverter.
@25°C: 4000W continuous + transfer (16 x 230 = 3680) = 7680W. However, your inverter will heat up, so the converter’s power will decrease.
With Excel, you could extrapolate a function of the form f(T)=aT+b.
And, more simply, for the converter alone, this gives: P(T)=−25T+4625
This means a power derating of 25W per °C.
Add your power transfer capacity and you have a fairly reliable estimate of your inverter’s capacity.
Stef_POLLAK
(sites-autonomes.fr & Team Victron Formation)
4
Il est bien ton petit graphique !
Je suis pas sûr que ce soit totalement linéaire…
Mais je n’ai jamais mesuré réellement le comportement d’un M+ ou d’un Quattro dans une telle configuration.
Peut-être @thierry_cortassa a-t-il des infos plus précises.
Hello,
Thank you for your promptness and the detailed answers.
Is the electronics temperature data unavailable? (I couldn’t find it on the VRM page or in the cloud data export.)
I took the opportunity to carry out a 24-hour analysis on a day with particularly high demand, to see whether consumption habits change in the future…
Indeed, the electronics’ temperature data isn’t available. However, you can still use the temperature probe supplied with your M+, which is primarily intended for the lead-acid battery.
But I’ve been thinking about something: beyond temperature derating, your Pylontech battery bank could also be playing a part. You should check the database to see whether there might be a correlation between the overload and a slightly low battery voltage.
And yes, Pylontech batteries have a low nominal voltage because they use 15 cells. This means that, to maintain a converted voltage (230 VAC), you’ll draw more current from the batteries, which quickly increases the converter’s temperature. The converter’s output then drops, so you have to draw more through AC-IN, but you’re limited — resulting in an overload.
In any case, this needs a bit more investigation; I’m just exploring the possibilities.
There’s a nasty power surge of almost 5,000 W. Has a particular appliance in the kitchen been switched on? Perhaps this time it coincided with another unusually high load, hence the alert.
Could you make your database extract for that day available? (In xlsx format, please — I have my “routines”.)
To see whether the battery doesn’t have the necessary response time (often due to its SOC), and the M+ doesn’t have enough time to compensate via AC-IN.
When you zoom in, you can see a consumption spike lasting several minutes. It goes from roughly 3 kW to over 8 kW. What is this appliance that consumes 5 kW for several minutes? The fan oven switching on?
In the middle of this period above 8 kW, is there a failure in generation? Almost 2 kW lost.
Perhaps you only get the alert when both happen at the same time: high consumption and a loss of generation.
I’ve looked at your data and have one additional question:
How is your grid code configured in the M+ settings?
You also say that you limited your input to 16 A, whereas I can see limits of 9 and 14 A in the shared data.
For some overload warnings, your battery voltage is low and you are also close to the discharge limit, even though the SOC is “high”, which indicates a significant current draw from the battery.
The Grid Code is “None” (feeding energy from DC to the grid not allowed)
Indeed, we reduce the input current during certain periods, so it is understandable that the values differ.
“With certain overload warnings, the battery voltage is low and you are also close to the discharge limit, even though the SOC is ‘high’, which indicates a high current draw from the battery.”
What hardware adaptations are required to eliminate this constraint?
Consumption can change quickly (cold room, industrial dishwasher, coffee machine, washing machine, etc.). We’re careful not to run too many appliances at once, but sometimes things slip through the net!!
Could you send me your configuration file (the .rvcs) so I can see whether we might be able to adapt the configuration rather than just tell you: change the battery…
Pylontech is, let’s say, meh, I’m not quite sure how to put it, but in any case, a pain, even though it works. This is due to their 15-cell architecture and the lack of active balancing. And when it comes to updating the firmware, they’re “a real pain” about providing the files. In short, Pylontech, eh.
