Townsend snow loss model- how to calculate snow events

[Areeba Turabi]

Hi all,

Snow events is a parameter for the townsend snow loss model. Does anyone have any recommendations for how to calculate snow events given snow fall data? How much time between active snow fall is a large enough break to consider them 2 separate snow events?

In case this is helpful, looking at how the townsend model uses these values, it seems like the more snow events you have, the less ground interference you have and the less "effective snow" is used in the calculation. Is there less ground interference with more  snow events because it is implied that there is melting between snow events?

Thanks,

Areeba

[Abed W Ayyad]

Hi Areeba, 

Trawling the internet, I found this: https://www.solaranywhere.com/support/solar-energy-modeling-services/snow-loss-modeling/ 

From what I can tell, the model they mention here assumes that snow slides off of tilted panels and that there is no allowance for anything other than sliding off or snow melt. Perhaps you can also account for some sort of realistic intervention on the human side? 

Again though, this is just me guesstimating and others on this list definitely know more!

Abed 

--
You received this message because you are subscribed to the Google Groups "pvlib-python" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pvlib-python...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/pvlib-python/6b133e86-d0ed-446f-9f4d-b3a7064b2e1dn%40googlegroups.com.

[Areeba Turabi]

Hi Abed,

Unfortunately this link is for the "NREL" snow loss model: https://pvlib-python.readthedocs.io/en/stable/reference/generated/pvlib.snow.dc_loss_nrel.html. But the model I am interested in is the "Townsend" model: https://pvlib-python.readthedocs.io/en/stable/reference/generated/pvlib.snow.loss_townsend.html. The question I had was specifically how to calculate the "snow events" parameter for the Townsend model. But thank you for your response! :)

Thanks,

Areeba

[Eliot Bryant]

Dear Areeba,

From my understanding, Townsend model classifies a snow event is an event where snowfall is greater than 1 inch.

At my company, we consider the number of days in a month where snowfall is greater than 1 inch (3cm) to be 1 snow event. So if 12 days/month had snowfall greater than 3 cm, then consider it as 12 events that month. It is true that this may not be 100% following the model as there could be 2 snow events in 1 day. But I have found that with this approach, the results of the Townsend model come within expectations and its results are the best of any model I have found.

Sorry I cannot help further, my suggestion is to direct the question directly to Tim or one of the other authors.

Warm Regards,

Eliot

--

[cwh...@sandia.gov]

>  From my understanding, Townsend model classifies a snow event is an event where snowfall is greater than 1 inch.

I've been through my correspondence with Mr. Townsend and the source paper, and I can't locate a specific definition of an snow event.

Snowfall > 1 in, or perhaps > 1cm, seems to me to be appropriate. A snow event should be enough snow to have an effect on energy production.

That's the best I can offer here.

Cliff

[Areeba Turabi]

Hi all,

Thanks for your input on how to calculate this parameter! In case it helps anyone in the future, I reached out to Mr. Townsend and got the following guidance (which is a similar approach to what you all have suggested- 1 inch threshold):

• An event has a 1" threshold.
• Since the std datalogging method is to record snowfall once per day (usually at midnight), the max events per day is one.

Thanks,

Areeba

[cwh...@sandia.gov]

It would be great to add this information to the pvlib.snow.townsend documentation, if you are willing to submit a pull request.

Cliff

[Areeba Turabi]

I'd be happy to submit a pull request! I'll get that out soon.

Thanks,

Areeba

[cwh...@sandia.gov]

Thanks. You can add text to the parameter description or use the Notes section.

Cliff