I’m pretty new to all this and trying to get a realistic handle on things.
So I’ve been trying out some things and must say I’m now a little confused…
But first…my installation 4x 385W panels MPPT 150|35 Multi 1200VA Pylon US2000C Cerbo
2 Days ago I was running my usual small load - fridge, desktop, wifi, router…still grid-tied, ESS and all that. As per screenshot attached below I then boiled a kettle off the fridge-adaptor to see what’s the what-what. This took about 3-4mins. My inverter (system in general) managed, no worries. As you can see the AC load was about 2300W for that short stretch. So that’s almost double the 1200VA.
Today I dropped the breaker on my DB that feeds AC into my installation to try an “Off-Grid” scenario. So far so good. But when we experimented, at 12h30 with a hairdryer (AC out on VRM showed 1222W) the whole system crashed, alarms, dead for 20-30secs.
Now we’ve used this hairdryer on this system already, albeit still when grid-tied, no problems…
I don’t understand these differences. The one day the inverter manages double. another day a bit over the 1200 margin and sniff sniff. What’s going on here? And how can a 1200VA manage 2300W for 3minutes?
Also, my system just cut out now again (12h48) for some odd reason, but for a second, as the fridge clicked on (116W), and then restored. No alarms. Have I hurt it somehow?
When you are connected to the grid, the difference comes from the grid. This is a common point of confusion for people, because they think of the output of the Multi is somehow “separate”. When the Multi is grid-connected (with ESS installed), the output of the Multi is literally connected to the input, and the inverter feeds additional energy (up to 900W for your 1200VA rated unit, because grid-connected is limited to 80% of total capacity) into this combined connection. That is why the kettle worked.
When you disconnected the AC, you overloaded the Multi. The Multi can go up to 200% of its rated power for about a second (to start large inductive loads), and it can go 20% over as long as the battery voltage doesn’t sag and it doesn’t overheat. Your 1200VA unit cannot power your 2kW kettle.
Your inverter can relieve your draw from the grid by up to 1200W.
It will not be able to sustain this indefinitely and will throttle back when it gets hot.
Exceeding its capability without grid assistance will shut it down.
A kettle is typically in the 2kW range, so a kettle alone would be sufficient to overload a 1200W inverter.
When a motor starts ( such as a fridge compressor) it draws many times more current than it normally runs at. Unlike a kettle, a motor draws a power surge to start which could also momentarily overload your inverter. The inverter has a very short 2x overload capability, but may not be long enough for a motor.
The long and the short of it is that your inverter is too small, and should only be used for very light loads such as lights, routers and low-power electronics.
Many people get by with a 3kW inverter and dedicate it to specific loads.
I personally find that a 6kW capable inverter is about the minimum if you want a hassle-free load-shedding experience and be generally able to use what you want with the minimum of overloads
Just another question:
In the screenshot below you’ll see on the top graph where I cut the AC.
The bottom graph shows the bumps where the fridge clicks on, then the ‘hairdryer’ spike, then one-and-a-half bumps, the second half being right now. With the two bumps before the spike the system did not switch off for a second or so, but now after the spike it does as the fridge clicks on. I know this because we also have a desktop with a UPS attached that now goes “CLICK-CLICK” as the fridge turns on.
Is there a chance I’ve somehow damaged the multi?
Or this what you might deem “Normal” behaviour? (as per @Phil.g00 's explanation)
I’m not sure I understand your question exactly. If you mean that the AC appears to momentarily dip just enough to upset a UPS on the same line, when the fridge starts, yes, that is perfectly normal. It is not really going “off”, but there is a momentary dip in voltage, which is unlikely to upset anything else.
That is simply because the fridge (I do not know the wattage) has a starting peak of around 5 times it’s nominal value, so it causes a power transient and the control loop in the inverter plays a little catch-up.
Perfectly normal. Trust me, one does not “simply” damage a Multi by overloading it. We periodically overload them during testing. In fact I saw one test where a 15kVA was repeatedly overloaded a few seconds apart for several days. The contactor used for the test later welded itself shut, and the Multi itself simply shut down and is still around…
I have a Bosch A++ fridge with a 90W compressor. These days I find that most new things have a compressor somewhere between 90W and 110W. That would give it a starting transient of 450W to 650W. As a percentage of your inverter’s capability, even my lower estimate is going to cause a slight hickup. It isn’t as noticeable on a 3kVA, but you do still see the odd flicker of a lamp.
Thank you @plonkster
That does set my mind at ease.
We have noticed that in some instances the fridge’s start-up is a bit higher, then the UPS squeals and whines for a second, but a desk light attached the same adaptor as the UPS, but not UPS protected, doesn’t dip or flicker, which I take as good news and evidence of your statements.
LED lamp I assume. They have switch mode power supplies and can generally deal with dips in the voltage quite well, to the extent that you don’t see them dip or flicker like an old incandescent lamp would.
But this prevalence of switch mode power supplies is exactly why it isn’t such a big deal these days. Almost everything has them now. I can use an example from my own living room. When the power fails, or a large load starts, the television doesn’t even blink, but the AV-receiver (a 15-year old Yamaha with an old school transformer) clicks off and back on. It is literally the only appliance in the house that shows any awareness of what just happened.
So I inserted an incandescent into the system and it does indeed dip.
Good news for me though, just ran an experiment where we included a washing machine into this system (off grid) and we managed fine; that’s a work-desk (wifi, router, UPS, laptop) / 100W globe / fridge / washing machine, all at midday to grab more PV. AC output shot to 2088W, then gradually sunk down to 800-odd W. Slight dip in globe as washy spin turned on. All good.
I’ve looked for this 80% in the documentation, but have not found it. Could you direct me, if you know where?
I watched a video of someone testing the grid power quality with an oscilloscope, in the USofA, seems their grid power is quite “dirty”. Once a Pure Sine Wave Inverter was attached between the grid and oscilloscope (thus, ultimately, one’s appliances) a lovely clean wave of current showed, so if the “…output of the Multi is literally connected to the input…” would this ‘cleaning up’ still be in effect or is it now negated?
So given these two quotes by @Phil.g00 and @plonkster above is the correct way of thinking about this Multi with ESS like zo: If the grid is a donkey and your load is a cart your inverter (&battery&PV) is a man standing behind the cart giving a little shove?
(I have questions about the ESS in general so may start a new thread…)
I have no freekin idea where that is documented. Also, this is something that can change in later firmware versions. You can verify this empirically too, of course. It is permanently limited to 80% when grid-coupled, and to 60% if it gets hot.
The reason is quite simple. The inverter rating (what it is advertised as) is actually higher than the maximum continuous power it can provide, and the reason for that is that in an off-grid setting, you will rarely use the peak power all of the time. One can have philosophical arguments about this practice, but the reality is that the entire industry does it. It also depends on the ambient temperature. The maximum continuous power is therefore limited to 80% (to prevent overheating).
Some of this, I think, is marketing. A poor grid connection CAN be helped by fitting an AC-coupled inverter.
Imagine for a moment a grid connection through an undersized cable. The voltage will be affected by large loads (because V=IR, and R is the problem here), but now imagine that you also have loads with non-linear current draw (in other words, the current draw is not sinusoidal). This is more typical than you might imagine. Cheap switch mode power supplies can easily draw all the power they need using only a portion of the sine wave. See here for an example, Figure 2 shows a nice example of non-linear current draw.
Now apply a non-linear current draw over a high impedance AC connection, and the voltage waveform is going to start showing the same noise.
Add a PV-inverter on that side, and magically the waveform improves, and marketing will quickly tell you this is because the grid is so much worse than an inverter. Which isn’t really true. A good grid connection is much cleaner than an inverter sine wave. A good large Diesel generator is better than an inverter sine wave.
Inverters often have a “steppy” sine wave. If you put a scope on it, and you drill in really far, you will see the wave is made of little square steps, not a smooth sine wave. A good grid connection does not have that. I can confirm that such a steppy sine wave is present with the Multi.
The multi is a current source when connected to the grid. It will not do anything good with the voltage waveform(except add some HF components, which is noise). To “fix” a poor sinewave, you have to do double conversion. That said, a sinewave that looks bad is probably not as bad is it looks - it’s more meaningful to look at it in the frequency domain.
The 80% thing is because the inverter is not actually 1200VA in any meaningful way, it can never deliver 1200W sustained. It can do 1200VA with a very specific load - in my opinion it is meaningless to write it on the inverter (except to dupe unsuspecting buyers).
A bit like writing 318 on the back of a BMW (while it has a 2.0 liter engine), and then slapping that same engine into the 328i (with a better tune). A lot of it is marketing. You cannot avoid marketing. It needs a “model number” of some sort.
The cheap Chinese inverters are even better at it. I just love those 5000W 12V units. I mean that’s like 400A… but don’t worry about that… just buy our inverter
From the Victron ESS design and installation manual: In a grid-parallel installation, the size of the inverter/charger can be (much?) smaller than the highest expected nominal and peak loads. For example, to cover the base load of a two-person household, an 800VA inverter/charger may be sufficient.
…reading through the thread’s comments above, and my general studies (though admittedly I was always the kid with the peashooter, not the one in the front row “Sir! Sir! Me sir”) I infer that I may be able to attach most of my house to my installation (whilst grid tied with ESS) and thus get more out of my PV.
Currently, even during my Off-Grid experiment all of last week, I was still only consuming/producing 1.6kWh - 1.9kWh / 24hrs out a possible 6kWh (4x 385W panels).
I should hope to even be able to attach my geyser (150L/3kW) and heat it during PV optimal hours 10am -15h00…tough this may overload my poor little 1200VA.
Is this feasible?
Of course, I should be able to disconnect most of these appliances when the grid fails. But it seems to me that if the magic of the ESS and the purpose thereof (partly) is to utilise PV and save on grid costs this little scheme makes sense.