How to Consider PV Production Losses from Snow Coverage?

When installing PV systems in areas that get snowfall, you should consider snow as a source of PV production losses. These losses can be significant during the winter months.

SolarNexus uses PVWatts and exposes all of PVWatts loss parameters for you use in modeling production. The PVWatts model provides a single annual loss percentage parameter for snow coverage losses. This very simplified approach seems to work fairly well for annual production predictions, however, it's not very good when trying to calculate hourly avoided costs from utilities.

Here are some known studies on the topic:

• Performance Modeling and Valuation of Snow Covered PV Systems: Examination of a Simplified Approach to Decrease Forecasting Error (2018)
• Integration, Validation, and Application of a Snow Coverage Model in SAM (2017)
• Measured and modeled photovoltaic system energy losses from snow for Colorado and Wisconsin locations (2013)

As these articles mention, the only available data that could potentially be used is old TMY2 data, and it can only be applied using a customized algorithm.

General Guidance

If your utility does annual true-up net metering AND the electric rates do not vary seasonally by very much, its sufficient to use a simple annual % loss for snow in your production estimations. A good source of these % loss estimates is this NREL document (see pages 20-21).

If electric rates vary substantially from season to season, you may want to try assuming snow as an addition to monthly shade losses. Each of the array areas has its own shade loss assumptions where you can input a higher percentage during the winter months.

NREL's study says "results from the model implemented in SAM should only be factored into annual considerations and not applied to monthly or shorter time periods." So if the overall percentage losses should be considered for annual numbers, we can targeted calculations of monthly percentage losses to just winter months such that when totaled, those winter month losses will equal same kWh loss given NREL's predicted annual loss.

For example, given:

• Site location = central MN.
• array tilt = 20 deg

So if we look at production without any snow losses for this representative system, we have:

Looking up approximate annual losses per the provided maps in the NREL study shows approximately 10% -16% annual production loss due to snow. (NOTE that steeper tilts reduce snow losses).

If we assume a 13% annual loss, that's about 1795 kWh of the 13809 kWh with no snow.

So if we distribute that lost production across 5 months (Nov - Mar) with this % distribution:

* Since SolarNexus automatically reduces shade by 33% when using MLPE's (shade mitigation factor, SMF), we would need to increase the shade number by 33% to get the same reduction in production.

If we input those % losses into the solar access %'s for the defined array area, it will significantly reduce the system's production during the winter months while leaving all the other months alone.