Recently I was thinking that it would be interesting to take a look at the amount of energy I use (and generate) to see how it varies over time. I already had daily data for my electrical and gas use, but thought that it would be interesting to see if this correlated to the weather. Luckily I found a weather station near my house that had a treasure trove of data, including solar radiation numbers. It took me awhile to pull all the data together, but I was able to come up with an interesting set of charts. I'm also trying something new in this post by embedding the charts directly, which should make them interactive. Let me know how it works for you!
First up is my gas usage. I figured it would be most interesting to try to correlate this with the average daily temperature. There seem to be two different, linear operating regions, divided around 55 degrees (F). Above this point my usage is relatively flat, while below there's a clear trend toward higher usage at lower temperatures. I use gas for hot water, home heating (hydronic), and cooking. My water heater is a Navien condensing tankless unit that I had installed in October of 2013.
Next is my electrical usage. There's not much of a trend here, though there seems to be a slight bias toward higher use at lower temperatures. This makes sense, since my electrical use is largely for lighting and appliances. My heating system only uses electricity to run small fans in wall units. Operating my clothes dryer causes a large spike in my electrical use, and is probably the cause of most of the higher values in this chart.
In order to examine the effect of the weather on my solar production it's useful to look at a couple of the weather trends in isolation first. The chart below shows the amount of energy that was available to my solar panels to collect each day. It's interesting to see the variability in this data even over short periods of time.
To help sort through the variations I also plotted the solar exposure versus the average daily temperature. This did not make as much of a difference as I'd hoped. It's interesting to see that the exposure amounts are grouped into two different bands for temperatures between 50 and 60 degrees. I wonder if this corresponds to Spring versus Fall patterns.
| Panels | Inverter | System | |
|---|---|---|---|
| Mean | 13.69% | 91.13% | 12.49% |
| Median | 13.95% | 93.49% | 12.68% |
| Std. Dev. | 2.03% | 6.31% | 1.92% |
| Maximum | 18.76% | 95.09% | 17.45% |
| Minimum | 1.08% | 31.03% | 0.34% |
My panels (Itek IT-255) are rated for a peak efficiency of 15.58%. Based on the 2014 data mine had a median efficiency of 13.95% with a standard deviation of 2.03%. My inverter is rated for a maximum efficiency of 97.2%, and had a median of 93.49% with a standard deviation of 6.31%. Overall, this put the efficiency of my whole system (from sunlight to usable electricity) at a median of 12.68% with a standard deviation of 1.92%. Clearly there's still a lot of room for improvement!
I know that the efficiency of solar panels is affected by temperature, so I wanted to see how mine fared. Excluding a handful of outliers on very cold days (the panels were covered in snow!) there's a pretty clear trend toward lower efficiency as the temperature increases, but there's quite a range of variability. Note that this chart is looking at panel efficiency rather than production, so solar exposure shouldn't be playing a large role here.
Nonetheless, I took a look at the panel efficiency compared to solar exposure. This shows what percentage of the available energy was converted into electricity versus the total energy available that day. Excluding the three snow days there's a downward trend as solar exposure increases. I hope that this can be chalked up to the correlation between high solar exposure and high temperatures, and the negative effect of those temperatures on the efficiency of the panels.
Last but not least I plotted panel production as a function of solar exposure along with a trend line. This shows that my panels will generate about 127 watt-hours of energy for each additional kilowatt-hour that falls on them.
Congratulations on making it all the way through! If you have any suggestions for other uses of this data I'd be glad to hear it.