Mixing modules of various types

[laurahin]

Hi,

With the state of module imports to the US, we're getting crazier and crazier mixes of modules in plants we need to model.  We're trying to keep the number of variants to a minimum and are wondering whether we can mix the following in the system    (Actually, we know some of these already but are just being thorough so this can be a reference)

- modules of different ratings, same manufacturer and series

- modules from different manufacturers, but same physical size

- modules from different manufacturers, different sizes

- multiple racking configurations, e.g., 1P and 2P

- fixed tilt and SAT racking

- bifacial modules with different bifaciality factors

- bifacial and monofacial modules

- bifacial and monofacial modules where we trick PVsyst by editing the monofacial module PAN to make it bifacial with a bifaciality factor of 0

To what extent would the shading scene need to accurately represent these, i.e., do all of the modules with different ratings need to be located in the correct positions in the shading scene with or without shading objects?  On the correct slopes? 

Thanks!

Laura

 

 

[laurahin]

I should clarify that I'm not just asking whether PVsyst will run, but whether the parameters associated with each subarray - IAM and bifaciality factors, power going into assigned inverters with corresponding ILRs, etc. -- will track through the system appropriately.  

Edited December 12, 2023 by laurahin

[Michele Oliosi]

Hi !

Except notably for

• irradiance transposition (by orientation)
• backside irradiance modeling (one bifacial model for the whole system in v7), and
• shadings if not using the module layout, (e.g., if “according to strings” or “linear” the shadings are calculated by orientation)

the calculation is mostly done separately, sub-array by sub-array.

I will try to summarize this as a table:

Mixing / Model	Note	3D scene without Module layout	3D scene with Module layout
modules of different ratings, same manufacturer and series	Defined as different sub-arrays	OK, but shadings are averaged over all modules in the same orientation	OK
modules from different manufacturers, but same physical size	Defined as different sub-arrays	OK*, but shadings are averaged over all modules in the same orientation *Bifaciality: not ok if rackings are not the same size	OK* *Bifaciality: not ok if rackings are not the same size
modules from different manufacturers, different sizes	Defined as different sub-arrays	OK, but shadings are averaged over all modules in the same orientation	OK
multiple racking configurations, e.g., 1P and 2P		OK* *Bifaciality: not ok if rackings are not the same size	OK* *Bifaciality: not ok if rackings are not the same size
fixed tilt and SAT racking	Not possible currently (→ v8)
bifacial modules with different bifaciality factors	Defined as different sub-arrays. Bifaciality factor is averaged between sub-arrays following an average weighted by DC nominal powers.	No particular issue with shadings since these do not impact backside irradiance modeling currently	No particular issue with shadings since these do not impact backside irradiance modeling currently
bifacial and monofacial modules	Not possible, unless bifacial model is deactivated
bifacial and monofacial modules, where we trick PVsyst by editing the monofacial module PAN to make it bifacial with a bifaciality factor of 0	Defined as different sub-arrays. Bifaciality factor is averaged between sub-arrays following an average weighted by DC nominal powers.	No particular issue with shadings since these do not impact backside irradiance modeling currently	No particular issue with shadings since these do not impact backside irradiance modeling currently

[dtarin]

The ability for PVsyst to model different portions of a plant in the 3D scene to different subarrays is possible, it's already there for fixed tilt systems, there are just restrictions in place preventing it from working for other systems. For example in fixed tilt, several orientations must be defined in the orientations menu and they must be unique orientations, otherwise it prevents you from moving forward. Computationally there is no reason why we couldnt define two orientations at the same tilt and azimuth, its just a restriction. But if you had two unique orientations in the orientations menu defined, then in the shading scene , you could allocate one or more portions of the plant to different subarrays. I presume that when doing so, the shading calculations will be specific to the modules defined in the orientation menu of the 3D scene and thus apply correctly to the subarrays when using according to strings. 

For tracker systems, this is not possible, but if the several orientations menu in the 3D scene was enabled and the user allowed to create orientations  with certain tables and allocate to different subarrays, then this would separate out the shading losses and be very convenient and moving the software forward to more realistically capture inverter level behavior to match operational data. 

Edited December 15, 2023 by dtarin

[dtarin]

If PVsyst, 

1. removed the restriction to allow identical orientations in the orientations for fixed tilt systems, and
2. allowed multiple orientations with tracker systems under orientation and enabled the option in the 3D scene,

then users could model subarrays with independent shading. If there are other restrictions in place for trackers such as with POA or bifacial calculations, then for tracker systems, a straightforward implementation would be to require all orientations to be of the same orientation.

[dtarin]

A monofacial fixed tilt test with two arrays with different ILR, orientation, and external shading had a difference of 0.05%, comparing the individual runs summed with a single combined run, which is reasonable (all on fast sim). With only a glance, it looks in the shading calculation for the 20/15 orientation there is a difference in the attenuation for diffuse and albedo in the combined run vs the individual run. 

Edited December 15, 2023 by dtarin

[dtarin]

42 minutes ago, dtarin said:

For tracker systems, this is not possible, but if the several orientations menu in the 3D scene was enabled and the user allowed to create orientations  with certain tables and allocate to different subarrays, then this would separate out the shading losses and be very convenient and moving the software forward to more realistically capture inverter level behavior to match operational data. 

This is worth highlighting. If PVsyst enabled this and users were able to define each inverter and their associated tables (or approximately close) as a unique subarray both electrically and with respect to 3D shading, then they each become an independent calculation path down the inverter level, and we could then get individual inverter outputs from PVsyst instead of a single combined figure, in addition to POA and other variables. There are some factors I am probably glossing over and other restrictions still in place (bifacial), but its likely better than a single model all mixed together with averages. 

 

[laurahin]

Thanks so much for the thorough response.  Just curious:  Did you try the option of converting a monofacial module to bifacial with a bifaciality factor of 0?  If not, we'll try it and I'll post the results.

[laurahin]

10 hours ago, Michele Oliosi said:

shadings are averaged over all modules in the same orientation

This implies that users should bear in mind that the shading animation is only for educational purposes.  

[dtarin]

I have not. As noted in the other thread, among the other considerations for bifacial, there is rear side IAM loss, which is not a loss that's provided but built into PVsyst, so there will invariably be a difference (and might be the largest contributor to the difference one sees). 

Edited December 15, 2023 by dtarin

[dtarin]

1 hour ago, dtarin said:

A monofacial fixed tilt test with two arrays with different ILR, orientation, and external shading had a difference of 0.05%, comparing the individual runs summed with a single combined run, which is reasonable (all on fast sim). With only a glance, it looks in the shading calculation for the 20/15 orientation there is a difference in the attenuation for diffuse and albedo in the combined run vs the individual run. 

The reason there was a difference in the attenuation is because I had trees present on the other array (20/0) to create a significant difference in the shading calculation to see how PVsyst compared in individual vs combined runs. With the trees removed, the attenuation in the combined run for the 20/15 orientation now matches the individual run of the same orientation.  So it is worth noting there will be a small deviation in shading calculations if external objects or irregular terrain are present (my assumption on the latter).

[laurahin]

On 12/14/2023 at 8:08 AM, Michele Oliosi said:

Except notably for

• irradiance transposition (by orientation)

Say you have an array with two different module types, same racking and orientation, but different module sizes.  Is the difference in backtracking behavior caused by the modules having different lengths taken into account when transposition is calculated for this array?  If it wasn't for the size difference/backtracking, the modules would all face the same way.  Thanks!

[Michele Oliosi]

In PVsyst v7, there is only a single tracker orientation at a time, i.e., all trackers move based on the same backtracking parameters. These are “table size”, “pitch”, and “top and bottom frames”.

If the difference in module size impacts the difference in tracker table size, then this may impact the backtracking parameters, depending on your choices in the “Backtracking management” window. However, ultimately, all trackers would still face the same way.

In PVsyst v8, you will have several tracker orientations, i.e., you can have some trackers move one way, and other trackers move another way.