I am working on a design with 24kw SMA tripower inverters and I was told by tech support that all 24kw can be installed on one MPPT as long as the input amps is below 33A. I have worked with ABB trio inverters in the past and I believe that the total was split between the MPPTs. The issue is when you have say 5 strings @ 6kw each. If you put 3 on one side and 2 on the other is the first side going to clip down to 12kw?
Designing for 2 MPPTs
- Thread starter coop3339
- Start date
- Status
jaggedben
Senior Member
- Location
- Northern California
- Occupation
- Solar and Energy Storage Installer
Practically speaking, that's probably only going to be possible with 1000V system voltage and the right panel specs. But sure.
No. That's the whole point and meaning of having a separate MPPT.
Of course the total AC output is going to clip at the rated output no matter what you do. So if you have 30kw facing the same direction in good sun you're going to get clipping, period. But if the 3 strings are facing east and the 2 strings are facing west, and it's morning, you should get the full benefit of the 18kW on the east facing MPPT.
Based on looking at the specs, it does seem that the limits on the Tripower might be more flexible than on some other inverters where you are forced to put a smaller percentage of the inverter's full potential on each MMPT.
Carultch
Senior Member
- Location
- Massachusetts
On Solectria inverters, they have an undocumented rule anticipating exactly this issue. The limit is that one of the 2 inputs to the inverter, can only be loaded up to 135% of half the inverter nameplate power, in order to avoid zone clipping. You will not find this documented in their manual. Their apps engineers told me it directly. The Solectria inverters come with a pair of jumpers, to tie the two input zones together, such that they each process half of the DC current for that particular inverter.
I am not aware of SMA or other manufacturers having this issue, of imbalanced sourcing of the two MPPT zones. The intent of providing the two MPPT zones is such that the two halves of the inverter can each be sourced by dissimilar array configurations. Ideally, the inverter's DC-to-DC converters for the two zones, should be built with a safety factor to handle the potential imbalance of power between the two zones.
The 33A limit on the SMA Sunny Tripowers refers to maximum usable current, which would be your total value for "Imp" per zone.
ggunn
PE (Electrical), NABCEP certified
- Location
- Austin, TX, USA
- Occupation
- Consulting Electrical Engineer - Photovoltaic Systems
But be advised that you cannot do this with the Tripower inverters. This is one downside of that design; you have to combine strings to go into the inverter, but you cannot combine into one pair of conductors so you are stuck with an 80% derate or two conduits if you combine on the roof. OTOH, with the Solectria inverters it is difficult to combine on the roof because the terminals you have to use to bypass the fuses are very small screws.
electrofelon
Senior Member
- Location
- Cherry Valley NY, Seattle, WA
- Occupation
- Electrician
But I'm confused. I thought the OP said that all the power CAN go into one input?. We almost had this question a little while ago. We had 6 strings and anothe guy started landing the strings from left to right in the combiner which meant 4 strings we on one mppt and 2 on the other. I caught it and assumed the power had to be balanced or close to it. We ended up balancing them for another reason so never had to investigate
ggunn
PE (Electrical), NABCEP certified
- Location
- Austin, TX, USA
- Occupation
- Consulting Electrical Engineer - Photovoltaic Systems
No, the OP said that SMA tech support told him that he could just use one input as long as the current didn't exceed the limit of the input board. What you cannot do with a TriPower (also according to SMA tech support) is tie the inputs together at the inverter like you can with a Solectria inverter.
electrofelon
Senior Member
- Location
- Cherry Valley NY, Seattle, WA
- Occupation
- Electrician
Ok so in other words one can run the entire inverter rating into one input IF the string voltage is high enough (about 720 vmp or 18 modules, generally)?
Carultch
Senior Member
- Location
- Massachusetts
Only if "all power" is less than what the Inverter manual says is the maximum usable input current per MPPT. The 33A for the Sunny Tripowers. Very common for the 12 kW size.
For the US version of the Tripowers, it is 33A for both zones. It is best if you can balance it as closely as possible.
For the rest of the world, it is extremely disproportional. 33A for zone A, and 11A for zone B. Even more frustrating is that you need to carry individual strings to the inverter, because each string has to enter thru a Sunclix connector.
Carultch
Senior Member
- Location
- Massachusetts
I'd be interested in understanding why you can tie the inputs together for the Solectria inverter, and why you cannot do the same for the SMA.
Probably has more to do with the formal listing, than any physics of how the inverter works.
In any case, to me it would be obvious that the jumper must have enough ampacity to carry the full current from either MPPT zone to the other, in the event that one of the DC-to-DC converters fails. The Solectria jumper is a factory tested part. To attempt to do this with the SMA inverters, the installer would have to be creative to find the terminations to do this with field-installed wires of proper ampacity. Jumping between the 30A unused DC Conn-U fuse blocks wouldn't be enough.
From SMA Data Sheet:
"Max. Input current / per MPP tracker ........66A/33A"
That indicates max 33 Adc per input. And 66 Adc only if you use both inputs.
---------------------------------------
So let's see..could we get 30 KW onto one input?
Assuming 60 cell, 250W, 9A, 40Voc modules.
That would be three, not four, strings per input, right?
Assuming the higher 1000 VDC design....and 40 Voc panels with an adder for colder than Cali temps...... That would be.....~ 22 panels per string, TOPS.
That would be 3 strings of 22 mod.s @ 250 W = 16.5 kw dc = max system size.
Snap.
------------------------------------------
What would happen if a single, combined dc circuit - of up to 66A - were wired across both inputs(AKA jumper)???
"Max. Input current / per MPP tracker ........66A/33A"
That indicates max 33 Adc per input. And 66 Adc only if you use both inputs.
---------------------------------------
So let's see..could we get 30 KW onto one input?
Assuming 60 cell, 250W, 9A, 40Voc modules.
That would be three, not four, strings per input, right?
Assuming the higher 1000 VDC design....and 40 Voc panels with an adder for colder than Cali temps...... That would be.....~ 22 panels per string, TOPS.
That would be 3 strings of 22 mod.s @ 250 W = 16.5 kw dc = max system size.
Snap.
------------------------------------------
What would happen if a single, combined dc circuit - of up to 66A - were wired across both inputs(AKA jumper)???
All this means is you have to run two dc circuits as OP mentioned, instead of one. Say...12 and 18 KW each.
Everything points in that direction..... there is a Connection Unit, as SMA calls it, a CB specifically made for this inverter. It forces you to use two output channels and can handle 8 inputs.... but only 4 per channel. And 33A output max per channel.
Everything points in that direction..... there is a Connection Unit, as SMA calls it, a CB specifically made for this inverter. It forces you to use two output channels and can handle 8 inputs.... but only 4 per channel. And 33A output max per channel.
Carultch
Senior Member
- Location
- Massachusetts
Good for people who don't have rapid shutdown requirements, and who can mount that combiner in a NEMA3R manner immediately adjacent to the inverter.
But not so much for people who have building systems, which could require a 3rd party rapid shutdown combiner, remotely locating it from the inverter, and laying it flat due to available rooftop space. In these cases you often have to be creative, since equipment rated for this application is slim pickings. SolarBOS makes such a combiner, and you'll need to find a 3rd party standalone DC disconnect if you need to remotely locate it from the inverter. Good luck finding one that has the appropriate voltage, amperage, DC application rating, and proper amount of usable blades.
I am aware of the locating the inverter within 10 ft of the array solution, and the 15 degree rack solution. Easier said than done, in many cases.
Carultch
Senior Member
- Location
- Massachusetts
Theoretically, it would divide the current evenly between the two MPPTs. They would each independently achieve the same voltage result, and process half the current. This is exactly the way you desire it to work.
In practice, you aren't permitted to do this because of the testing, listing, and labeling of the equipment. Small nuances in the effective resistance of each MPPT zone as a load, could make the current divide disproportionately.
ggunn
PE (Electrical), NABCEP certified
- Location
- Austin, TX, USA
- Occupation
- Consulting Electrical Engineer - Photovoltaic Systems
ggunn
PE (Electrical), NABCEP certified
- Location
- Austin, TX, USA
- Occupation
- Consulting Electrical Engineer - Photovoltaic Systems
jaggedben
Senior Member
- Location
- Northern California
- Occupation
- Solar and Energy Storage Installer
Which is what I said in post #2. (BTW I said 1000V system voltage, by which I meant up to 1000V. Hope that was obvious).
And it also may depend on the module short current spec, because the allowable 43A is borderline between 4 and 5 strings for typical modules nowadays.
No, it probably has a lot to do with the MPPT algorithms. The MPPT inputs might very well work against each other and not be capable of dividing the current evenly in a stable manner.
On ABB inverters (formerly Power-One) there is a 'parallel' vs 'independent' switch that changes the behavior for the two inputs. If the inverter is set to 'parallel' but the strings aren't physically paralleled on the array side of the inputs, the inverter will sense this and throw an error. Each input has a power limit so it's sometimes necessary to use the parallel mode if you have an odd number of strings.
So presumably SMA could design such a setting into their inverters ... but they haven't.
Which is good enough reason for me.
- Status