In Architecture 244 (The Secret Life of Buildings), the first task we were assigned was to predict, measure, and analyze the thermal conditions of our home refrigerator over the course of 24 hours (midnight to midnight).  In class, we were given a paper with a grid, and were asked to estimate the fluctuations of temperature that may occur.

As one might expect, we all came up with diverging estimates – from temperature range to grid scale to peaks and slopes of the temperature line. My estimate is shown in green in Figure 1 below – my assumptions were that the temperature of the refrigerator would remain relatively constant, only to fluctuate when in use (i.e. when the door would open).

We were provided a HOBO temperature data logger and programmed the device to record the temperature of the refrigerator every 5 minutes. I placed the device in the volumetric center of my refrigerator (on my housemate’s shelf) at 11pm on August 31^st and left it to record data through September 1^st. It was moved out of the center shelf at some point during the course of the day, presumably so that a housemate could access her container of spinach. I am not sure (and she does not clearly recall) at what point of the day this occurred, so it remains unclear how and whether this affected the results.

The HOBO data somewhat aligns with my initial assumption as an overall trend of the temperature raising when the door is opened (see 7am, 12:30pm, 6pm, which approximately align with meal preparation times), but also illustrated that the temperature fluctuates on its own within 2 degrees as the compressor cycles on and off and creates a sawtooth curve-like behavior. I did not expect quite this level of constant fluctuation in my estimate.

Because the interior our house tends to heat up during the day (and the refrigerator is in a south-facing room), I explored the relationship of the effect of sunlight and the temperature of the refrigerator by overlaying a graph showing the height of the sun above the horizon across the same 24-hour period (based on longitude & latitude data, specific to September 1, 2010, via a calculator created by the University of Oregon Solar Radiation Monitoring Laboratory). I represented the height of the sun instead of the average exterior temperature in an effort to more easily visualize a) the fact that this graph represents a 24-hour period, from midnight to midnight, and b) to indirectly represent the change in temperature of the exterior environment. The thermal relationship between the exterior environment and the refrigerator isn’t clearly apparent in the results, but is still worth consideration and comparison to others’ results.

The remaining issue I am unclear on is why the setpoint temperature seemed to have dropped from the morning to the evening: between 0:00 and 12:00, the setpoint appeared to be around 34 degrees, but after 2pm it seemed to be at 33. This may be because the location of the HOBO changed during the day towards the back of the refrigerator, which may have a colder average temperature due to it being further away from leaks. This is my hypothesis of why the setpoint appeared to change, but I cannot confirm this without additional research (perhaps two HOBOs recording temperature at the same time but in two different locations.)

Figure 1: Thermal conditions of my refrigerator, estimated (green) and actual (red), and the elevation of the sun on September 1, 2010

This entry was posted on Wednesday, September 8th, 2010 at 5:30 am and is filed under Refrigerator Data. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.