No, I am not more in contact with chinese supplier since quiet a while, because they are too expensive.
I mean this really serious.
I pay for Trina Dolar 425Wp only 15,1¢/Wp including VAT.
If I import from chins a 40ft container I have to pay transport, TARIC and VAT and then I am at 22-23¢/Wp.
Hence, I simply go to my local Electric Distributor (NOT RESELLER) because I am a professional installer in Estonia.
However, my 425Wp Trina Solar are 1130x1780mm and this is already 2m2, which is already horrible.
I could have gotten another brand with 650Wp, but they have 1300x2300mm hence 3m2 and you need a whole team to install this monsters.
How many wattpeak does your 1000x2500 or 2000x2500 (I have never seen such size) have?
The dimensions are also important to me because of the limited roof space.
For 1000 x 2500 mm, I can get 550-600w (144 cells). I’m trying to keep voltage and current manageable.
Same range for 2000 x 2500mm. That size is weird. I have never seen it before either. Apparently, they use 120 cells. I guess the cells are bigger than standard.
I am not quite sure what you are saying. On the 450/100 the two trackers are, as you say, effectively independent and both seem to accept the maximum from each string evenly (sun permitting). In my case that is 3kw/string as the strings are not connected but simply (and nothing more than) eight parallel panels in series. This gives me 6kW peak. I suspect that on occasion the input power exceeds 6kW as the output power occasionally has and there is the ~8% loss to contend with. I rely on Victron being decent designers and will have overtemperature protection and over current/voltage/power protection and the whole thing is solid state and computer controlled. If you have panels capable of delivering over 5kW then the 450/200 is fine or two 450/100 if you value backup. Requires a very simple array in four sections. I admit in winter I would probably get maybe 15% extra output with a four-array setup due shading.
What is your total panel output?
I accept and take your point that I have no redundancy, I would have to hope local suppliers have stock for fast dispatch. OTOH a day’s worth of night electricity (to batteries) only costs £0.90 so the cost of not having backup for a few days is not huge.
I suppose we should have asked, waaay up at the top, how many panels and how much output you are aiming for?
I ask because you are talking about using a 250/50 which only outputs ….. ooops I can only find a 250/60, is that the one? Assuming that is so, it’s only delivering max 3.5kW and you have two, so OK 7kW max.
What are your individual panel outputs and how many?
OK I am guessing you have 10 panels arranged in a pair of 5-in-series. I am guessing these have a s/c current of 14.4A (as stated above) and are probably something like a JA 445W panel with a VOC(max) 39V. Expected max output 4.5kW, VOC 200V. That’s too much power even for a single 250/70. Yes you should have diodes, loads of these dirt cheap I would get 40 or 50A.
2-off 250/60 or 1-off 450/100 will do. Problem solved, but ultimately your installer decides.
Yes 2x 250/60 looks a good choice.
Ah, not quite sure how this will work. I haven’t looked into it (I have no need) but you will likely have to make one a master and the other a slave as I am not sure a cerbo can manage multiple chargers. You have not told us the whole setup. Is it PV - MPPT - 48V Batt - inveter - mains, all controlled by a cerbo? You may need to add a smartshunt in/out of the battery as the victron stuff is inaccurate, partly because it doesn’t handle energy efficiency losses well (programming of cerbo/inverter not great here).
@farmeroz This is purely a guess, I stand to be corrected.
I was thinking that:
Victron cannot guarantee that shorting will succeed when the MPPT short circuit threshold is exceeded.
By the time the MPPT takes shorting action, some current will have passed through. Your MPPT is likely protected if your panels are below the short circuit threshold. Victron cannot guarantee that if your panels’ current is above the MPPT’s short circuit rating.
I didn’t include all the details because I assumed they were not relevant and my thinking wasn’t crystalized. I have since decided how to move forward.
I will install 24 panels, 8 parallel strings, of 3 panels in series each string, on 2 MPPTs. Contrary to the examples I gave before, I will use two 250/100 MPPTs, with each having 4 strings. My panels will have 560W, 50.20v VOC and 14.24 ISC each. Everything will be symmetrical. So I will stay well under the voltage and short circuit limits of the MPPTs but over-panel by 137% on both MPPTs (this is ok by me). The MPPTs will charge a bank of 48v batteries of 900Ah/3072KwH with internal BMS. This installation will be in the tropics, so plenty of power.
The connections will be straight forward. PV→string combiners → MPPTs (connected to Cerbo GX) → busbars →batteries. The busbars will connect to a Quattro 15KVA, which will also be connected to the Cerbo. I will use the Quattro’s 2 AC-Ins to connect to the grid and the Genny. All loads will be on the AC Outs of the Quattro.
I won’t parallel the MPPTs as this is not vital. I will enable DVCC, with CVL and CCL. Yes, the Cerbo is able to coordinate the charging (of all connected Victron MPPTs and the Quattro). As a backup, I will program all charging equipment with charging parameters.
with 24 of yor 560W 50V panels, you can put 4 panels series in a string, than parallel
three strings into one MPPT 250/100 TR or
two strings into one MPPT 250/70.
As an alternative you could use an RS 450/200 with 20 panels, which gives same output like the two MPPT 250/100. The RS 450 will max to 200A at any temperatur wiht 4x5 panels.
The remeining 4 panels can be connecte to an MPPT 250/60. Thouhg:
1x RS 450/200 : 4 string with 5 panels
1x MPPT 250/60 : 1 string with 4 panels
If you have different directions, than one RS 450/200 with four strings up 7 panels are benefical. Each singel tracker can deliver up to 80A according Victron MPPT Calculator. Sum of all 4 MPPTs cannot go above 200A. Though with east / west directions, the 4x 6 panels RS 450/200 can beneficial.
If redundacne is key, I would go with 3x MPPT 250/70 each with 4 panels per string and two strings in parall. This eleminates the back current diodes, as needed only wiht 3 or more strings in parallel.
That’s a good option. If I recall correctly, over-paneling might be even higher but that’s ok with me. I will check more closely. As I already have the two 250/100, I could add another 250/70 or even 250/85.
Anyway each 250/60 has a short circuit current (max) of 30A (per MC4) and you are only delivering 14.4 so you are well in. One string of five cells in parallel (<<245V) max short current 14.7A (<<30) is well in spec.
OK its 24 off 560W 50V panels so short current probably about 12A and Ptot about 14kW.
so 8 strings of 3 through four 250/70 ‘s gives 150V x 24A/device each delivering 3.5kW looks just about ideal to me.
Well in on everything. Note that in the datasheet for the 450/100 the PV short circuit current is defined as REVERSE polarity, as another poster noted.
I’m seriously considering going with your suggestion @BjoernK of 3 parallel strings of 4 series panels each running to each 250/100 instead of 4p3s. However, I can’t find any reputable reference to the effect that with 4 strings, I will definitely need diodes but with 3 strings, I don’t need diodes. Could you please provide a reference or explain the theory?
If you have two strings parallel, you do not need string diodes. If string A has less voltage than string B, string B cannot harm string A.
If you have 3 or more strings parallel, you need string diodes.
If string A has less voltage than strings B and C, than strings B and C can push their combined current to string A, which is more than string A can handle.
The higher the string voltage the less are losses.
Surely, solar cells are a reverse biassed diodes. As such, applying a voltage below the breakdown voltage of the diode will not result in any reverse current. Most of the current will be going in the opposite direction due to the photoelectric effect. The maximum reverse voltage seems to be called the “max system voltage” as seems to be in the region of 700-1000V for commercial panels. There is thus no effective limit (other than limited by the inverter) for having any number of single equal strings in parallel. For parallel connections, the differences in panel output will balance out, losing some efficiency as most will not be operating at the optimal current/voltage but will average out. That can be easily inferred from the graph below. What you don’t want to do is forward bias one panel from the output of another, but offhand I can only see that happening in more complex setups with panels in arrays of combined parallel and series combinations. Or I am missing something?
@farmeroz I also assumed those were reverse biased diodes but my vendor insists that they are bypass diodes. Maybe they perform both functions in this case? Maybe my vendor is wrong?
It’s hard to see what use diodes would be in the scenario I suggest (ie parallel single strings) I set above. I presume the reverse breakdown voltage of the panel is ~1kV, which is far in excess of all common inverters in domestic use. So the panels do not need overvoltage protection (save for lightning strike). Each string is simply a voltage source with characteristics as above/below.
It’s unclear to me what happens above 55V in the example above, but from first principles (half remembered from 50 years ago) it’s not impossible that current will flow negatively against the inbuilt junction potential due to free optocarriers crossing the junction. I can find no graphs that cross into negative PV current potentially forced by imposed voltage exceeding the highest PV voltage attainable, so I suspect it is in fact just a reverse diode above that with close to zero current as per the graphs. Below is a more useful graph, perhaps. Actually, thinking about it from distant memory doing the quantum mechanics of junctions I think this is the case. Someone here with a spare PV cell could do the actual experiment, assuming we see the sun any time soon!
Hmmm, @farmeroz I assumed that any extra power would be cut off provided the ISC, VOC and short circuit current limits were observed? Isn’t this what over-paneling is about? For what it is worth, the online calculator recommends the smaller 250/85?
Anyhow , for this particular project, I have moved on to two 450/100s because of other considerations. But for a future project, I still plan to use two 250/100s.
Well, I would hope the devices have over-power protection and would automatically limit their power input and thus output, but generally it’s not a great idea to deliberately provide significantly more power than the devices are designed to handle, IMHO (its in the spec). I am not sure we know what power output is allowed by the ISC. I am allowed 10kVA for example. Yes, a lot to be said for 450/100’s but these are rated at 6kW each (I have one), so better, but not actually in spec. I have two 8-strings but lower output panels than yours that rate out at 5.96kWmax but do on occasion slightly exceed this. Overpanelling is good, but in my experience of limited use. I am either close to full output during a sunny day, or well down due clouds. On sunny days I am exporting 100% of my solar pretty quickly.
