Challenges in Simulating Shaded Photovoltaic IV Curves with Boost Converter in Typhoon HIL

Question

Good afternoon, I’m Igor from UFSC.

We were testing some curves for the Typhoon HIL with a boost converter and, when testing the same thing as in other simulators, we noticed that the curves with shading, generated via CSV, have a certain "problem" when fixing the value of D, where the voltage becomes fixed.

First, we tested with .ipv files, and the voltage never went above 37V, even after changing the maximum voltage and other parameters of the curve. So, although the curve was drawn as we intended (Image 1), it didn’t work as expected and the voltages were limited.

Then we tested other possibilities to circumvent the issue, trying to run with .ipvx curves, with curves without shading (Image 2), and it worked perfectly. With the Duty cycle at 0.5 — meaning with an output of 400V it should remain at 200V — it was perfectly stable and surpassed the "limit" of the .ipv.

However, as this wasn’t our goal, we wanted to obtain a curve with shading from 2 modules. We tested with a manually created curve using data from the CSV (Image 3). Even though it wasn’t very accurate, it reached much higher voltages. However, when setting the desired voltage (400V) and applying the curve with D=0.5, the voltage varied significantly — it didn’t remain at 200V, but fluctuated between 150-250V.

I thought it could be due to the curve being a Normalized IV, but another test with a curve without shading worked perfectly.

So we found a temporary solution: we added more than one module to obtain shading in the Boost (Image 4). It worked, but it wasn’t possible to properly visualize it to generate .pdf, .png, or any other type of graph that characterizes the curve and its behavior, like we had with "Photovoltaic" in Scada. We need this for research article purposes.

So our questions are, is there any way to:

1. Generate a .ipvx curve with shading using the Waveform Generator? (in a way that works accurately.

2. Make the "Normalized IV" curve I generate via CSV more accurate?

3. Plot a graph for the solution of n modules in series?

4. Generate a .ipv with CSV that works for a wider voltage range?

The answer to any of these questions could help us improve our work.
Thank you in advance!

Answer 1

Hi Igor,

The Normalized IV curve type is better suited for voltage and current scaling, because you can specify the values of open-circuit voltage Voc and short-circuit current Isc. The .ipv file format is more suitable in cases where you do not need to modify the data set imported from the CSV file.

In the Image 3 of a Normalized IV curve that you shared, the IV characteristic, which is colored in blue, is not monotonically decreasing. There are multiple voltage values that correspond to a single current value. Maybe that is the cause of the voltage oscillations you observed.

Perhaps you specified the voltage and power data sets in the CSV file instead of the voltage and current data sets, because the shape of the IV curve in Image 3 usually corresponds to a power-voltage curve with shading. The first column is your CSV file is used for voltage data points, and the second column is used for current data points.

For the solution with n modules, there is no support available to show a single IV curve for multiple modules connected in series using the preview functionality in SCADA.

Please find attached an example of a CSV file and the corresponding Normalized IV file for a shaded curve.

photovoltaic_panel_shaded.csv

photovoltaic_panel_shaded.ipvx