Inverter Losses

[pplatt]

Do you have more descriptive definitions of the particular inverter losses? I was looking in various PVsyst literature and was unable to find more in-depth definitions of the following losses:

InvLoss Global inverter loss

IL Oper Inverter Loss during operation (efficiency curve)

IL Pmin Inverter Loss due to power threshold'

IL Pmax Inverter Loss due to power overcharging

IL Vmin Inverter Loss due to low voltage MPP window

IL Vmax Inverter Loss due to upper voltage MPP window

Thank you.

[André Mermoud]

InvLoss Global inverter loss

- This is the sum of all inverter losses.

IL Oper Inverter Loss during operation (efficiency curve)

- The inefficiency loss, computed according to the efficiency curve.

IL Pmin Inverter Loss due to power threshold'

- Loss when the power of the array is not sufficient for starting the inverter.

IL Pmax Inverter Loss due to power overcharging

- When the MPP power is over the input power required for obtaining the specified PNom(ac), the inverter displaces the operating point on the I/V curve in order to get exactly the required power for Pnom(ac). This IL Pmax loss value represents the difference between the Pmpp and this adjusted power.

NB: The displacement is towards higher voltages. If the voltage exceeds the Vmaxmpp limit of the inverter, the inverter has to stop, and the Pmpp is fully lost.

IL Vmin Inverter Loss due to low voltage MPP window

IL Vmax Inverter Loss due to upper voltage MPP window

- If the Vmpp is outside of the inverter's window (Vmppmin/VmppMax) the inverter will clip it to the limit value. This loss is the difference between Pmpp and the corresponding P of the I/V curve at the limit value.

[cheinzer]

THank you for the reply. It was very helpful in evaluating the .csv file. I have evaluated several of the files and discovered that the inverter Pnom or nameplate max AC output is often surpassed on the output file. In this case, it was for 11% of the operating hours, with the total amount of time the inverter was clipping 27% of operating hours. When clipping was not occuring though it was supposed to, the inverter loss due to efficiency was negligible, and the inverter loss due to max power was 0. This made the inverter look very efficient on the output page.

The file doesn't stop at max power, but efficiency is right

 

On a different simulation however, the inverter clipped appropriately and the Pmax losses were correct, however now the inverter losses due to efficiency were relatively high vs. the previous simulation, where if the inverter was producing at max nameplate, the efficiency losses were zero, which makes sense. I would argue once the inverter is at nameplate, all the losses should be due to power threshold. Taking it one step past that, I would even argue as is seen in this screen shot below, that if the inverter is producing at 2kW below nameplate (in this case, 18 500kW inverters) or 7998, the max efficiency loss should be 2kW and the rest should go to power threshold.

 

clipping modeled correctly, but efficiency confusing

 

Can you please explain why the difference and how that happens?

Thanks in advance. I am also Swiss, but from the German part living in the US. I hope to soon make it back for one of your trainings.

Edited November 1, 2012 by cheinzer

[cheinzer]

I have also noticed that there are InvLoss when there is no power production, and it almost seems random. Any idea what may be causing this? I have attached a screen shot of my simulation output .csv to show this.

[André Mermoud]

I don't have definitive answers. Which version are you using ?

Until version 5.57, there were an error in the simulation process: when the Vmpp was lower than VmppMin (il_VMin loss), the overpower was not checked.

This may have some effect in systems with low Vmpp value (by hot conditions).

Since 5.58, this has been corrected: The overpower increases the Voper value, and is comptabilised first. Usually this brings the Voper over the VmppMin value.

Now with highly oversized arrays and high Vmpp (more modules in series), you have the risk that the overpower correction leads to an operating voltage higher than VmppMax, and in this case the inverter has to stop completely (it cannot find a suitable operating point), leading to high overpower losses.

The losses at very low irradiance are quite normal, I don't see where is the problem.

[scrussell]

A slightly different question, but could you please explain the following terms in the Loss diagram?

 

• Inverter Loss during operation (efficiency)
  
• Inverter Loss over nominal inv. power
  
• Inverter Loss due to power threshold
  
• Inverter Loss over nominal inv. voltage
  
• Inverter Loss due to voltage threshold

 

It's the second one ("Inverter Loss over nominal inv. power") that I'm having most trouble understanding.

thanks,

Stewart

[André Mermoud]

Did you read the first answer to this post ?

What is still unclear ?

[André Mermoud]

You are right.

This is explained in detail in the FAQ Why sometimes the oberload losses increase significantly without reason ?