What We Have Learned

Dual-axis tracking systems increase electricity production by about 30% when compared to fixed panels

A picture is the best way to illustrate this point.

Tracking PV Daily Graph Non-Tracking PV Array

In this example, the tracking system produced 32% more energy than the fixed array on a normalized basis. The dual-axis tracker essentially makes the day longer and sunnier. It always faces the sun at the ideal tilt angle (high tilt in the winter and low tilt in the summer) and azimuth (east to west). This allows the system to produce about 30% more energy!

The non-tracking array will only be perfectly optimized for a short time each year. This forces the system owner to pick the best tilt angle and azimuth for the array.

Panels at a 25-degree tilt angle produce about 5% more electricity than panels tilted at 45 degrees

The answer to this lies in the old farmer's adage to "make hay while the sun shines." Essentially, our 25-degree tilt arrays are better optimized for the longer, sunnier summer days than our 45-degree tilt arrays. This allows the 25-degree tilt systems to slightly outperform the 45-degree tilt systems over the course of a year.

Snow can help or hurt system performance

We have some arrays at a fixed tilt angle. As a result, snow cover on the panels can cause the systems to lose several days of production during the winter. The snow generally melts off during the first day of temperatures of 32 degrees or greater after the snowfall.

This graph shows the sunlight and electrical output on one of our arrays the day after a snowfall. You can see that although it was a very sunny day, we did not produce meaningful energy until after the snow melted off shortly after noon.

Monona Grove

On the other hand, we have seen snow cover on the ground serve as a reflector for the sun and actually increase the output of our pole-mounted tracking arrays. The tracking array does a good job of shedding the snow in winter because it is at a steeper tilt due to the sun being low in the sky during the winter months.

The Lussier Heritage Center

This graph shows the reflective effect of snow on the ground in front of one of our pole-mounted Solar Sunflower tracking arrays. The sunlight exceeds the value expected on a clear, sunny day (1,000 W/m2). As a result, our array that is rated at 1.1 kW-AC produced 1.4 kW-AC, a 27% increase!

Monitoring can identify design, equipment or site problems

The vast majority of solar PV systems have been installed without a comprehensive monitoring system. Since MGE is actually researching solar PVs, each of our systems has, at a minimum, a utility-grade electrical meter that is read every month. Most of our systems also record weather and solar information.

Our comprehensive monitoring packages allow us to identify design, equipment and site problems that cause our PV output to fall below expectations.

  • We found a design problem when we noticed one system was performing particularly poorly compared to our other systems. The culprit was a large transformer specified by the inverter manufacturer that remained energized throughout the night. Even when power is not flowing through transformers, they continue to consume energy and are a parasitic load for any PV system where they are part of the design.
  • We installed a relay to de-energize the transformer each night and re-energize it when the sun came up. It is now controlled by a photo-eye and only allowed to operate during daytime hours. As a result, we have cut our parasitic losses by over 80%, and PV production is much more acceptable.
  • We found an equipment problem with a new German-made inverter. This popular inverter was brought to the U.S. market with the same output rating as in Germany. U.S. laws are more restrictive about how hot equipment is allowed to become. To comply, the inverter manufacturer installed a derate program on units for sale in the U.S. This causes it to produce less energy when it gets hot.


  • We noticed on sunny days, although we would get a nice curve of sunshine, our electrical output curve had a nasty notch in it right around noon. The notch lasted several hours while output was reduced (see graph). We brought this to the manufacturer's attention, and they provided us with cooling fans at no charge to keep the inverters within required parameters. As a result, the inverter company has changed its literature and now does not recommend installing their inverter in closed indoor spaces without the new $90 fan. Without real-time data monitoring, we never would have noticed the notch in the output of the inverter.
  • We found a site problem after analyzing data for one of our first systems. Again, this system was underperforming, and we started to investigate. The culprit here is a very common one. Our site has large fir trees on both the east and west. As a result, our array takes longer to get going in the morning and stops producing sooner in the evening. This system is experiencing shorter days than normal and, as a result, produces about 27% less energy than an unshaded system.

For more information