Sunny April

I've mentioned before that our Sunpower systems here at SMC seem to be significantly out-producing the production estimates created by the NREL web-based tool called PV Watts. One never knows if this is the better Sunpower technology or merely more available sunshine than normal. When we see that the systems are out-producing the estimates almost every month, we think that the Sunpower claims may well be valid, and we're looking for data from other manufacturers' systems to understand this better.

However, April 2012 was just plain sunny as all get-out - no April showers here. The PV Watts April estimate for an optimally oriented 4.76 kW array is 538 kWh. In April, our system made 802.5 kWh! Here's the graph of actual output vs. estimate:

In the depth of winter we have shading from trees to the south, so making less than the optimized estimate makes sense. It's interesting to see August and September being so close, and then how things shift in 2012. There's been a drought here, and we need more rain, yet if we have to have a drought, it's nice to have a clear benefit as well.

 

Comments

[Eric]

What pvwatts parameters did you use - in particular did you tweak the AC->DC derate factor at all?

[Marc Rosenbaum]

This model was done with the default derate factor of 0.77.

[Erik Haugsjaa]

Using PVWatts 1 for Worcester, MA with 0.77 derate and our Canadian Solar 230W PVs and Enphase M190s, we have a similar 2012. (10% high so far) VS our total for 2011 was 99.6% of the PVWatts estimate.

All very inexact since we have some shading that I am not factoring in and the M190s clip occasionally in fall and spring. I suppose the low 0.77 is acting as our shading.

[Eric]

I did a similar post last week too ;) http://sandeen.net/wordpress/energy/comparing-pvwatts-estimates-to-actual-solar-production/ but used the 0.817 derate Enphase recommends. It's hard to tease out snow vs. shade vs. m190 clipping etc, but in the end PVWatts was reasonably close - I'm not overproducing like you are, though....

[Marc Rosenbaum]

Eric, with a 0.82 derate factor, using Boston weather data, the system is still out-performing the estimate, but not by as much of course - about 7.8%.