Running Hot and Cold

People in Miami can now feel slightly less guilty for using air-conditioning in the summer. According to a new study, cooling down hot cities in the US is far less energy-intensive than warming up cold cities.

There’s no question that AC eats up a lot of energy. But “the energy demand from living in a cold climate such as the Midwest or Northeast is frequently taken for granted,” writes the study author, Michael Sivak at the University of Michigan Transportation Research Institute in Ann Arbor, Michigan.

Sivak decided to compare Miami, the hottest major US city, to Minneapolis, the coldest. First, he calculated how often and by how much the cities’ temperatures strayed from 18 degrees Celsius. He also accounted for the efficiency of furnaces, boilers, air conditioners, and power plants.

The amount of annual heating that Minneapolis requires is higher than the annual cooling that Miami requires, Sivak reports in Environmental Research Letters. And that heating takes more energy than cooling, since furnaces and boilers are only about a quarter as efficient as central air conditioners.

Overall, Minneapolis’ energy requirements for climate control are 3.5 times higher than Miami’s, Sivak estimates. He speculates that since people “are generally more tolerant of heat than of cold,” Miami’s energy usage might be even lower than his study suggests. Roberta Kwok | 29 March 2013

Source: Sivak, M. 2013. Air conditioning versus heating: climate control is more energy demanding in Minneapolis than in Miami. Environmental Research Letters doi: 10.1088/1748-9326/8/1/014050.

Image © juan ignacio laboa |



  • Eli's Air Conditioning April 1, 2013 at 1:25 pm

    Why would people feel guilty for using air conditioning in the first place? Just to conserve energy? At the expense of their comfort, happiness, and potentially health? Come on.


  • Paul D April 3, 2013 at 6:28 am

    Interesting analysis. The assumption has to be made that all heating in northern cities has to come from fossil fuels to make this comparison work. If a northern city were to adopt solar thermal hot water and space heating along with geothermal or other district heating, the numbers would change. AC is almost exclusively powered by electricity, so how that electricity is generated (coal, gas, oil, nuclear, hydro, etc) affects the analysis. It’s not as simple as just calculating the end use energy consumption.


    • roberta April 6, 2013 at 9:33 am

      Great point. The researcher based this analysis on current sources of energy for heating and cooling in the US. He writes: “The heating sources in the US are natural gas (69% of energy generated), oil and other petroleum liquids (17%), liquid petroleum gas (7%), and electricity (7%)… The source of energy for cooling appliances is almost exclusively electricity… In the US, electricity is generated mostly from burning fossil fuels such as coal (42%), natural gas (25%), and oil and other petroleum liquids (1%); additional sources are nuclear (19%), hydroelectric (8%), and other (5%).” He then accounts for how efficiently power plants convert those energy sources into electricity.


  • Aaron D June 12, 2013 at 1:56 pm

    Although interesting, this article contains poorly constructed science tying multiple assumptions and averages together. National heating sources shouldn’t be used for actual heating sources in either of the study sites. As stated elsewhere, energy use statistics by region/state/city are available through the USEPA if you really want to look at energy use trends. This article adds little new knowledge and shouldn’t be featured in a quality publication. FYI, I was also disappointed that the chart displayed on page 9 didn’t even list what unit was being shown. Try a bit harder next time.


    • ldoermann June 12, 2013 at 2:30 pm

      The chart displays an index that combines annual “heating degree days” and “cooling degree days”. The authors give no units and explain it this way:
      “Heating degree days is an index of the energy demand to heat buildings. This index is calculated by subtracting the mean daily temperature from 18 °C (65 °F), and summing up only positive values over a fixed period, such as an entire year. An analogous index for the energy demand for cooling is represented by cooling degree days.” For more, see the original paper here:


Leave a Comment