Estimation of AC losses with micro-inverters

[julmou]

Good morning,

I have a system with 6x modules, each with an Enphase micro-inverter. I am aware of the fact that micro-inverters tend to increase the AC losses (mentioned in this post: https://forum.pvsyst.com/topic/1929-what-can-i-gain-with-micro-inverters/)

• First, maybe a suggestion, it would be nice to have the Mismatch Losses automatically down to 0% when selecting micro-inverters (or to zero when ticking the Default option), just like it does when selecting power optimizers, and perhaps also the DC losses down to zero or close to zero.
• But mainly, I want to know how I can estimate my AC losses with micro-inverters. It doesn't seem the tool in Detailed Losses is correct in that case. Example below:

I have 6x modules 500Wc. Now, obviously with micro-inverters, each module is in parallel with the other (each module being a individual string) and a single AC cable (with 230Vac) goes along all modules, and each module adds up its current to it -> at the end the AC losses should consider the total current in that AC cable. What it does instead is consider each inverter with its own cable separately. See the screenshot below. You can see that the estimated power in each "inverter to injection" cable is 0.48 kW, the power of a single module. Where in reality, all inverters/modules connect to a single AC cable, which thus carries more current and suffers from higher losses than if it was 6 cables with a smaller current => how can we estimate correctly the AC losses in that situation ?

Thanks a lot for the help.

Julien

[Michele Oliosi]

I see it is not very practical at the moment.

At the end of the day, the losses are ohmic losses, so a workaround could be to increase the length of the cable (here you could multiply by 6) to account for the higher current in reality (loss ~ RI^2). By computing the effective resistance of your cables you may get to a more accurate workaround.

[dtarin]

Calculate your voltage drops for each leg and the combined conductor to the point of interconnect. Set your loss fraction @ stc appropriately based on this. 

[André Mermoud]

For the Mismatch loss, I have indeed implemented that for the next version 7.2.17.  The mismatch will be null for one module in series, and half the default value for 2, 3, or 4 modules in series. Either for optimizers inputs, and for any other subarray  (i.e. Micro-inverter)

For the Input circuit resistance, it is not straightforward to decide that the inverter is directly mounted on the PV module when there is one, two or three modules in series. .  Therefore we let the user define a null-resistance if this is the case. 

 

For the interconnexion of several micro-inverters in a single bus, the problem is not quite simple. 

If you decide a same section for the whole bus, and a distance Lseg  between inverters: let us define  Rseg = the resistance of one segment between two inverters.  The loss will be    Ploss =  Rseg *  ( I² + (2*I)² + (3*I)² + ... )  etc. =  Rseg  *  I²  *  (1 + 4 + 9 + 16 + ... ).

Now in the PVsyst definitions, you have one wire per inverter, the total loss is to be accounted for the average length. Therefore the total loss is   PLoss = Ninv * Raver * I²  

Finally:      Raver * NInv = RSeg * (1 + 4 + 9 + ... + NInv²), from which you can easily calculate Raver to be put in the PVsyst  tool.  

NB: as  R = Length/Section, and Section is constant, you can replace R by Length in the previous expression:  

Laver = Lseg * (1 + 4 + 9 + ... + NInv²) / NInv