Adding an input for a reset value to the soiling models for cementation effects

[jdbeltman]

Is your feature request related to a problem? Please describe. I have been trying to compare soiling model methods and was hoping to include cementation effects where a rain event doesn't completely clean the modules in that analysis. Currently, the two models, hsu and kimber, both completely clean the modules at every rain event over the set threshold.

Describe the solution you'd like A simple floating point input from 0.0-1.0, representing the soiling ratio/loss that the model will reset to after a rainfall.

Describe alternatives you've considered I've so far used some post processing techniques to get a general idea, but I don't think they are rigorous enough.

[mikofski]

I wonder if experimenting with this line

https://github.com/pvlib/pvlib-python/blob/46851d91bb70e9b5521feb7ec119b057ec3d7f70/pvlib/soiling.py#L203

would have an effect? I was thinking make cleaning equal to some user input fraction of the soiling, that way later when it subtracts the “cleaning” it only cleans by the fraction?

[mikofski]

Also, do we need Adie’s permission to change her model like this? Are you planning to publish some results? I think it would be interesting

[cwhanse]

I like the idea of a parameter for the amount of soiling NOT washed off and can see how it would be both physically meaningful and useful. Like you, I don't see an easy way to achieve this effect using the existing parameters.

But pvlib also strives to provide stable implementations of published models. A way forward would be as @mikofski suggests: is there some document that can be referenced that describes this modification?

[jdbeltman]

Ah okay, apologies, I am new to using pvlib so I wasn't clear on what the rules were for requesting changes to the library, but it certainly makes sense to stick to only published models. I haven't at this time come across any published predictive models using soiling reset values. Thank you, Mark, for the advice on how to edit the code for myself.

[mikofski]

I would really like to see what you come up with and hopefully show your work at a PVPMC workshop or some other forum like PVSC-52 in Seattle. If you’re looking for data maybe try partnering with Groundwork or Fracsun or the Regional Test Centers run by Laurie Burnham from DOE. Please don’t be discouraged to contribute to pvlib.

[adriesse]

I would like to echo Mark's encouragement!