v6.32: Has inverter output power limit changed?

[jfish]

Andre,

Was there a change from 6.31 to 6.32 in how the simulation treats inverter output real power limit?

I just upgraded and had to go back to an earlier variant that I had modeled in 6.31. When I re-ran the simulation I saw that the peak output of the plant was significantly increased from the 6.31 simulation even though I had not changed any parameters or components. I went back to my v6.31 PVsyst version (I did a parallel install which I recommend to anyone) and saw the values return to the original outputs one I re-ran the simulation in 6.31. The only loss I see that changed in the waterfall chart was "Inv loss over nominal power". Since I am modeling an inverter that has different "nominal" and "maximum" values in the OND file I am wondering if you know are using the "maximum" values in the simulations or if there is any other simple explanation for this behavior?

[André Mermoud]

You are probably using a SMA central inverter.

The manufacturer has changed the specifications about the nominal output power in the database. It is now depending on the ambient temperature, and comprised between PNom and PMax.

[jfish]

Andre,

Yes I am using an SMA inverter but I don't see how this change came from the manufacturer. It looks like this change was from v6.31 to 6.32 only. I am using the same variant file and settings and same custom OND file between my two simulations using v6.31 and v6.32. There is no change made by me to the variant or OND files between the two simulations. In 6.31 I get clipping to Pnom and in 6.32 I do not. I realize that the SMA SC inverter can be set to adjust Pmax based on ambient temperature but prior to PVsyst v6.32 PVsyst used the value Pnom to clip inverter output, the Pmax setting in the OND file did not appear to do anything. There were no other settings available in inverter OND files for ambient temperature adjustment prior to v6.30.

It appears that you have now changed the programming of PVsyst and the format of OND files so that adjustments to PV output power is now made? Is that correct?

[André Mermoud]

I don't understand. I didn't change anything in the program concerning the Inverter behavior for the version 6.32.

Please send your full project (using "Files" / "Export projects") at support@pvsyst.com

[jfish]

I'm happy to do that Andre but I think I am getting a better understanding of the changes, though I am still confused how these changes were implemented.

I went back and looked at the variants and inverter databases for these two simulations, one in v6.3.1 and one in v6.3.2. I have gotten in the habit of always doing a parallel install which I find very helpful when upgrading since it helps me in better understanding the changes made in each new version.

I confirmed that no changes were made to the variant file between the two simulations. The change appears to only be in the inverter database entry and was not done by me.

I am attaching two images of the the inverter I modeled from the inverter database. One is when accessing the database in v6.3.1 and one is when accessing the database in v6.3.2.

The inverter entry in the database from v6.3.1 shows the ability to limit output power based on temperature but it does not have an adjustment variable set below 50C:

 

The same inverter entry in the database from v6.3.2 shows that a temp adjustment variable has been set for down to 25C:

 

I did not make this change to this database component. It appears to have been made during the upgrade.

If this was a change to the PVsyst component database, shouldn't a new database item have been created, keeping my old version unchanged?

Perhaps this is really a question about how component database items are handled during version upgrades. I was always under the impression that component files were unique and static once I import them, unless they are overwritten.

So in this case can you help me understand this process? Was there a change to this inverter definition between 6.31 and 6.32 and if yes, how does this change occur? Does PVsyst overwrite old database items during upgrades? If the answer is yes, can this be changed to give users an option? I would have expected that the user would be given the option to accept/reject any changes to these files since a change to an inverter affects all previous simulations using that inverter, which is in general, not desirable.

Also on this subject is there a way that in the PVsyst report you can make a note that inverter output to ambient temperature adjustment has been used?

I want to make clear that I appreciate these changes in principal since the ability to model different inverter performance at different ambient temperatures within PVsyst is a great new feature. I am only trying to better understand how changes are implemented and what I can do as a user to identify changes and prevent them if I do not want them made automatically.

Thank you,

[André Mermoud]

As I mentioned in my first answer, the Manufacturer SMA has indeed changed this in the database for the version 6.32.

Now we update the database using the requirements of the manufacturers.

They very often ask for modifying some parameters of existing devices. We don't maintain an historical logbook of all the modifications. With 10'000 PV modules and 3'000 inverters, this would be a very tedious work, which would double or triple the volume of these databases (and seriously complexify its maintenance, which is already a big task).

By the way even "official" Datasheets are never definitive: it is almost always mentioned that they may be changed at any time.

I have indeed not yet taken the time of mentioning the Temperature dependance of the inverter in the report. Sorry. I should do that for the next version.

[jfish]

Thank you for the clarification Andre.

Might I suggest that you make clear in your help documentation and in your other support locations such as: http://forum.pvsyst.com/viewtopic.php?f=19&t=1752 that database items in the PVsyst component database may be overwritten without any warning dialogue during version upgrade? Additionally I think it would be important to make sure users understand that there is no indication that these files are changed when they are overwritten during an upgrade and these changes can impact previous models that used this component definition.

I understand the complexity of maintaining this sizable component database and want to say that the addition of this database to the PVsyst program does increase the value that your software provides. My suggestion for you as you continue to develop this program is to develop a new policy when manufacturer's request changes to their component, or for any other reason that components are changed in your database, that a new file is added and a new version number attached to the file name. The old file should stay unchanged. The "data source" value is a handy place to place a date stamp for the new version identifier if you want to keep the same model number for the component.

My concern here is not that manufacturer's change their specifications, I understand that and agree that that is not your responsibility. My concern and the reason for my suggestions is that these changes appear to happen automatically during a PVsyst upgrade and that those changes happen without any indication to the user that they occurred.