Temperature log: main features
This graph, for 2016, shows a winter pattern of indoor and outdoor temperatures that is typical for this house. Indoor temperatures vary much less than outdoor temperatures, they rise and fall with them, and they are higher nearly all the time.
While the outdoor temperatures shown go as low as minus three degrees, those indoors lie within the winter “comfort zone” from 17 to 24 degrees (see this post) nearly all the time.
Weather this winter
This winter was harsh for a solar-passive house. Near-record rainfall (227 mm) came with the greatest number of cloudy days of any winter in the new century. There were 53 mornings with more than four octas of cloud, when the average is 33.
Because cloud limited the the solar gain, I had to use blower heaters far more than in previous winters. My records show that I used 320 kWh ($80) in these heaters this winter, when I normally use about 40 kWh ($10).
Heaters were also used by guests who were present on the six days shown. As well as being unused to the climate, the guests lived in the colder west wing of the house. They may have used 72 kWh ($18). Those guests have kindly written reviews of their visit.
Even using 400 kWh of electricity for personal heating in a winter could not make a detectable change in house temperature. I have found that blower heaters are surprisingly good at making a room in this house comfortable. As the radiant temperature of the walls is only 2 or 3 degrees too low for comfort, it can be compensated by making the air temperature only slightly higher.
The pattern in detail
While cloudy days are not plotted here (Cloud observations for this winter are plotted elsewhere.), cloudy days can be recognised on the graph. In this climate, days with low maximum temperature and high minimum temperature are always due to cloud. Only in fine weather are days warm and nights frosty. The graph shows how the weather goes through a cycle every week or two: sunny days get warmer, then rain sets in. As it clears, the air gets even colder, before warming up again.
Indoor temperatures follow the same cycle, but there are differences. There may be a delay of up to a day, and sometimes longer.
I did scatter plots comparing all the variables shown in the first graph and I fitted linear regressions. I present the four scatter-plots that had the highest coefficients of determination (“R-squared”).
The linear trends and means shown are like those that have recurred from year to year. However, they show the effect of this particularly cloudy season.
(1). Indoor maximum regressed on outdoor maximum
The indoor maximum temperature rose and fell only one quarter as much as the outdoor maximum did. On the average, the indoor maximum was still 3.1 degrees warmer than the outdoor maximum, even in this cloudy winter.
(2). Indoor mean regressed on outdoor mean
The indoor mean temperature rose and fell only one third as much as the outdoor mean did. As an over-all average, the house, at nineteen degrees, was about eight degrees warmer than the environs this winter.
(3). Indoor minimum regressed on outdoor minimum
The indoor minimum temperature rose and fell only 22% as much as the outdoor minimum did. On the average, at the coldest point of the daily cycle, the house maintained 17.7 degrees, which was over 13 degrees warmer than outdoors.
(4). Indoor daily temperature range regressed on outdoor daily temperature range
I have simplified this regression to a ratio: the indoor daily range is one fifth of the outdoor daily range. The mean daily temperature range outdoors was 12.8 degrees; much narrower than the usual winter value (15.4 degrees) due to the persistent cloud cover. Indoors the range was kept down to only 2.6 degrees by storage in the “heat bank” of bricks, concrete, and under-slab soil.