My electricity supplier allows me to enter regularly my power meter reading to monitor the energy usage over time. I did this 87 times during more than 2 years and wanted to better understand the trend.
It looks like the yearly trend in the period 2009 - 2012 is increasing:
 |
| Yearly usage |
whereas it seems like the weekly average was decreasing in the past two years:
 |
| Historical |
The decreases was expected because:
- heating pump exchanged with energy efficient model (Feb-13)
- stand-by killer (Mar-13)
- exchange lamps with energy efficient models (when they failed)
- two adults (Aug-13 ..Dec-13 and Feb-14 .. Aug-14)
Interpretation is a bit difficult, the following two aspects have a significant impact:
- holidays
- seasonal effects (daylight duration, temperature)
In order to better interpret this data I wanted to compare monthly averages with the previous year. Fortunately the data can be exported to *.csv format . Unfortunately Excel can't handle irregular time series data - fortunately with R's module zoo this is possible:
The R code snippets to achieve this will follow in the next blog.
No comments:
Post a Comment