In June, my colleagues and I took a tour of SAS‘s new 1 Megawatt solar farm in Cary, NC.  Standing in stark contrast to the high-tech metal/glass/silicon is a decidedly low-tech landscaping crew that keeps the grass around the panels trimmed to a manageable height.

Landscaping Crew

A flock of Dorper sheep occupy the pen with the solar array to eat the grass.  Sheep – because goats would jump on the panels – and dorpers because they’re short enough to walk under the panels and all of the associated electrical equipment.

As this is not an agriculture or animal husbandry blog, I should probably get back to my point.

SAS makes statistical software for a variety of applications, so it stands to reason that they would apply some of their number know-how to their own data coming from the solar array.  Back in May, Xan Gregg from the JMP software division, shared a visualization showing solar power output over the course of a day.  He asked “Why is there a small drop in power output around mid-day?”

See it in the January – April data?  Well, the real answer has something to do with the angle of the sun at various times of year and how the panels tilt on their single axis, but several people pointed out that heat can affect solar panel efficiency.  A hot panel cannot make as much energy as a cooler one.

To measure the effect of heat on power generation, Xan has come up with some new visualizations to answer the question.  Rather than steal his thunder, I’ll just provide the link and let him explain it in his own words.