You've reached step 4 of 6 in our 'Getting Started with Utility Scale Solar Farm Forecasting' series


No two solar farms are alike!

We understand that each solar farm is unique. Local topography may shade your solar farm in the morning hours. Soiling in arid regions is common.  Wiring losses across large solar farms are substantial.  Not all inverters are made equal.  

That's why we've created Solcast’s PV Tuning Technology.  In a sense, it is (almost) magic.  It learns the real characteristics of your PV plant, using your measured power output data and our solar irradiance and weather data.  It's a novel technology, which can make achieving your solar mission much easier.  

Once you've created your utility scale solar farm site, and learned how to access the solar farm forecasting data, it is time to provide us with an historical extract of power output from your solar farm site, so we that we can apply PV tuning to improve the accuracy of your forecasts. If you're a commercial user evaluating our service, we'll happily tune a single test site for you. We won't share your data with anyone else, and if you want to delete your measurements from our system at any time you can. Tuning will help, provided your site has been running at full capacity for at least 3 months, otherwise the standard un-tuned forecast will be best for your site.

Enable PV Tuning via a Historical Data Upload

Just send us the recent history of your site-total PV power output measurements.

Ideal data to send us:
1 minute data spanning the past 12 months until the current day, including power, availability, and constraints/curtailments. This will give you the best results. 

Minimum data we can tune with:
Power only data, at maximum 15 minute intervals, for at least the past 3 months. If your data is coarser than 1 minute, you'll need to tell us whether the values are instantaneous, or period-averages.

Data format to use:
You'll need to use the CSV format like these sample files:

Download ideal data sample file
Download minimum data sample file

As you'll note in the Ideal sample, if your solar farm asset has been operating at below 100% of its capability for any reason (constraint, maintenance, partial construction), you'll need to provide an 'availability' field.  This is a 0-100 % value which represents how much of the solar farm is currently online.   On example calculation for this field is

% availability = total number of inverters in operation/total number of inverters installed

Another would be as solar farm with an AC capacity of 50 MW, constrained to 30MW.

% availability = 30MW/50MW

How much data is needed?

At the minimum, we'll require at least two months of (recent) data.  Ideally, a full year of measurement data should be provided. 

Once uploaded, your site will automatically be updated so that the forecasts and estimated actuals are 'tuned' to capture the individual behaviours of your solar farm. 

You can download an example file of PV power output data for tuning here!

PV Tuning in Action at a Utility Scale Solar Farm Site

Pictured here is a sample of our utility scale solar farm forecasting product, including live accuracy reporting as available via the Solcast API. The user can visually compare the SCADA actuals (orange) to the Solcast satellite-derived estimated actuals (blue line) and the Solcast forecasts (white grey line). Once you've created your own site, you'll have access to live demonstration charts just like the one above.

Need more help? 

If the above information hasn't answer your question, please contact us via the chat bubble at right & we'll reply as soon as we can! --->

You can also consult our documentation or send us an email

Next Article: Step 5 of 6, Automate POSTing of solar farm measurements

Did this answer your question?