## Tuesday, December 14, 2010

### Anticipating my electric bills based on mean outside temperatures.

For all of December, I have been recording the temperature in three major rooms of my house (Living room, Dining area and Master bedroom) three times a day in order to get a feel for the average temperature that is maintained 24/7. I have had the thermostat set between 69F and 70F for the entire period. The end result of all this measuring has been to determine that the average temperature for the core of my home has been 68F.The central unit used to heat my home is a heat pump. There are also a couple of small space heaters that are used as needed.

Since I now knew what the average level was inside and had also recorded the average outside temperature from day to day, I was able to construct the graph pictured here. On the left side is power use expressed as kilowatt hours while the X axis displays what I call ‘load’ which is merely the difference of the average inside temperatures (68F) minus the average outside temperature for any given day. I then graphed each day as a point in a spreadsheet expressed as an x-y scatter diagram. To this set of data, I have applied a trend line that is extend backwards twenty units. As you can see, there is pretty close agreement when comparing the kilowatt hours used to daily average temperature (the R-squared value was 0.9101). This graph then, gives me a pretty handy tool for anticipating not only the cost of electricity for a particular day but also for a week or a month.

As I have data for the first thirteen days, so then do I also have a pretty good way to guess that this month will end up with an average temperature of somewhere between 32 and 37 degrees Fahrenheit ( the fifty percent statistical spread around the historical mean temperature of 35F that is left for the remainder of the month). Therefore, at ten cents approximately a kilowatt hour for electricity, I can assume a month end bill of between \$186 (60KWh x 31) and \$211 (68 x 31). An interesting side note to this experiment is the points that fall either above or below the trend line. I will assume that a point above represents heavier than normal use of appliances while those below, less. This is due to the fact that when measuring the power consumed for each day, I had to measure the total power used by all the electrical devices in the house, not just the furnace. Hope this made some sense.