After four weeks without HVAC, including eight days of no power at all due to Hurricane Helene’s impact on Georgia’s electrical grid in September 2024, I was reminded of an emphatic declaration that the late George Baird made about me nearly thirty years ago. He was introducing me for a lecture at Harvard, where I was junior faculty at the time, and he stated that “Michelle Addington doesn’t believe we need air conditioning in buildings.” At that time, I was thinking how much I wished I could go home to air conditioning and not to the stuffy apartment that I was renting in Boston. While George was basically on the right track, recalling that my research into micro-cooling was the antithesis of centralized AC, he perhaps wasn’t aware that I was focusing on alternative methods to the inefficient air handler–based systems whose use was becoming standard in much of the world, as I knew full well that the need for cooling was only going to increase.
In my teaching, I had been trying to pivot students away from the need to design for heating and direct them instead to focus on how we must design for dumping heat out of buildings, a.k.a. cooling. Heating is easy, it is what nature wants to do, whereas cooling is difficult. Convincing students and practitioners that they needed to let go of their celebrated methods for passive heating—exemplified and repetitively illustrated by Le Corbusier’s seminal “l’ete/l’hiver” section diagram of the Unite d'Habitation that encouraged solar admittance during much of the year as a means to provide heating—was met with stubborn resistance. It frustrated me that no amount of scientific analysis could shake them from an opinion about how things worked that traced back to a single image produced by a non-scientist nearly a century before. This image, which shows the ideal depth of an overhang or brise-soleil as determined by the angle of the sun at noon during the summer and winter solstices, presumes that direct winter sun is the ideal. But direct sun is notoriously problematic to use for passive heating, which is why there are no indigenous passive design strategies utilizing the sun, even in the coldest climates. Conserving and utilizing the heat we already generate on the interior has always been the primary approach.
When the oil furnace failed in the apartment I was renting in Boston during a bitterly cold February, I finally had a chance to demonstrate how much heat we readily produce. For two weeks, I was without heat, having turned down an offer from my landlord to move in with his family. I wanted to show how the first and second laws of thermodynamics could serve as a guide. I purposely did not use any space heaters, and instead moved my refrigerator and all equipment, especially the lamps, to the innermost part of the apartment, far from windows and the perimeter. I hung up sheets of butcher paper in front of—not on—the coolest surfaces to prevent mean radiant temperature loss, and set up a cot and makeshift desk next to the refrigerator’s condenser coils. The temperature in my “zone” never dropped below 65 degrees Fahrenheit (18 degrees Celsius). I can’t say that I wasn’t absolutely elated when the new furnace arrived, as showers in the colder bathroom were not fun, but I knew that I finally had a relatable narrative for translating the laws of physics into tangible and meaningful evidence for students.
Cooling is a different animal altogether, as it brings into play the third law of thermodynamics, which is particularly unforgiving. Although the majority of places I have lived in adulthood were without central air conditioning, they were all north of my childhood homes in the Washington, DC area, with the exception of my first apartment in southern Alabama. That apartment, along with my subsequent residences in Philadelphia, Boston, and New Haven in the US, as well as Rome, Delft, and Munich in Europe, either had no AC or were served with a single window unit or fan. Conversely, my homes in Northern Virginia, New Orleans, Nashville, the lower Delmarva Peninsula, Austin, and now Savannah have all been cooled with progressively sophisticated centralized systems. I don’t love that I depend on these systems and am so devoted to them in this humid part of the deep south. But this is also not to say that I wasn’t miserable during the summer in the northern latitudes, where I often found it too hot to sleep. I well remember the brutally hot summer of 1994 in Europe when the only way I could sleep in my stifling unit in Delft was by taking a cold shower every few hours, but I can’t imagine even attempting to live without central AC in the south.
I have always done my best to minimize my energy use. Lights are only turned on if absolutely needed and only when I am in that space. All major appliances are used only during off-peak hours, and even then, with discretion. And I primarily cool my house at night, so as to do my part in preventing the grid from having to fire up the dirtiest plants to meet demand. Nevertheless, the extended power outage and an even lengthier loss of HVAC in September 2024 pushed me to test how long I could manage. The first two days of no power were not a problem; I had precooled the house in anticipation of an outage and I had plenty of battery-operated fans and lamps. I had also packed my fridge-freezer with various phase change materials, a habit I started in Austin due to the surprisingly frequent power outages in the downtown district there. Ice is our most common phase change material—it holds temperature at a steady 32 degrees Fahrenheit (0 degrees Celsius) as it melts—and so is ideal in a refrigerator, whereas salt hydrates can be engineered to melt at much lower temperatures, thereby good for stabilizing freezer temperatures.
Two days after the hurricane knocked out power to three million residents in Georgia, I received a text at 4 am from the power company announcing that my power was restored and I quickly plugged in all my rechargeable devices and went back to bed. It was not until I woke up later that I discovered they had erroneously mis-phased our section of the grid, resulting in a larger, more complex failure, and more problematically, blowing out every DC converter that was plugged in, as well as much circuitry. I was not only back to being without power, but I no longer had any of my rechargeable fans and lights, as all had their batteries destroyed. Stores had already been cleaned out of rechargeable equipment and online retailers were projecting delivery dates weeks ahead. So I had to become more assiduous about preventing heat from coming in. I covered every possible aperture that sunlight could enter with curtains I found in the attic and the privacy shutters I installed when I purchased the home. I added further layers to a vestibule with sheets so I could enter and leave without admitting the hot and humid air. I tucked partially frozen bottles of water into my bed to cool it down (the phase change materials had kept the refrigerator cold for five days and the freezer at proper temperature for two, and at a low enough temperature to keep bottles of water partially frozen for nearly a week). And I spent as much time outside as I could so that my body heat wouldn’t contribute to the internal heat gain.
By the time my power was correctly restored, the house was up to 80 degrees Fahrenheit (27 degrees Celsius). Not bad when about half of the eight days were in the low-to-mid 90 degrees Fahrenheit (>32 degrees Celsius) with high humidity. Unfortunately, that is also when I discovered that the mis-phasing had damaged my heat pump and air handlers. Supply chain disruptions resulted in a predicted repair date of “not any time soon.” Fortunately, however, night time temperatures were beginning to drop below 70 degrees Fahrenheit (21 degrees Celsius), giving me an opportunity to start opening windows.
This is where I am particularly lucky. The homes in my section of Savannah’s historic district typically do not have window screens. I get it—they are ugly. I had debated removing them when I bought my house, but my engineer half beat out the designer half and the screens stayed. While I had opened the windows now and then on a lovely day, I had never really exploited the use of natural ventilation to exhaust heat during times of high temperature coupled with high humidity. When we think about natural ventilation as a means to move heat through spaces, we typically think of temperature as the determinant. This is a reasonable proxy in high pressure zones where the relative humidity tends to be low, and is what occurs during cold winters in most locations as well as summer months in dry regions. Low pressure zones, however, such as the tropics and the southeastern sides of continents above the equator, have high relative humidity. In these contexts, it is not temperature that becomes the primary determinant of heat, but enthalpy: the measure of the total heat content of air including temperature and the thermal energy of water vapor.1 Strategies abound for the passive cooling of high temperature/low humidity air, but all strategies for the passive cooling of warm temperature/high humidity air require either the direct removal of the water vapor or dilution with lower enthalpy air. Not so easy—but it’s the only way. So I spent the next three weeks measuring temperature and humidity throughout the house to take advantage of every tiny enthalpy difference to move heat. I constantly opened and closed particular sets of windows, as well as opening, closing, or angling shades depending on the location of the sun. In full disclosure, I made a sixty-mile drive north to snag the last available box fan in the region to facilitate dilution.
I had to pay attention, as one forgotten window shade or window left open on the wrong side of the house could ruin a night of cooling, but I found my rhythm, and I must say that the interior was not just pleasant, it was refreshing. Granted, by the time the air handlers and heat pump were operational again, we were finally out of the brutal heat and humidity of the summer, so none of this should have been too hard at that point. But the experience made me realize that I had been treating my central system as the default condition. I pledged to go forward with the default condition being systems off, and to only strategically use them to manage a steady thermal inertia, as well as to shave the peaks and valleys of the thermal conditions. When external conditions are either higher or lower than desired conditions, inertia becomes our greatest force. Thermal inertia is most dependent on mass, so high density materials used in large quantities as with masonry, stone, and concrete will drive conditions for low density/low mass air. I added an infrared thermometer to my tool box so I could measure surface temperatures of a concrete floor and other high mass construction materials. I spent the next six weeks testing this approach, using relative humidity and interior surface temperature as triggers. When either the humidity or the concrete floor radiant temperature started to creep up, I ran the system for an hour or two, or I opened the windows if the exterior enthalpy was lower than that of the interior. This was all I needed to maintain comfort.
I do not share the above narrative to gain sympathy—far from it, as I was both lucky enough to have escaped the terrible wrath the storm wreaked on so many, and privileged enough that I had plenty of options. It was my choice to stay in the house to test things out. Having spent the first half my career as a mechanical engineer who specialized in fluid mechanics and heat transfer and having applied this knowledge in the design and operation of thermal processes and power plants, I already had the toolkit and the ready knowledge to act quickly and decisively. Too many do not have such choices and opportunities, and many more, particularly those in the low-pressure zones across our planet, will be subjected to lengthening spans of the higher heat/humidity conditions already impacting millions as the Earth warms.
In the beginning of the sustainability movement, when many practitioners were trying out physical modifications—passive solutions—in search of a means to reduce the dependence on mechanical systems, their actions were often dismissed as ineffective or even “greenwashing.” Today, sustainability tends to mean high technology, and we have seen major efficiency improvements in lighting and in HVAC systems themselves. Low- and no-carbon sources of electricity can now be easily incorporated, but they don’t require us to rethink why, what, and how we design. Nor are they the product of architectural knowledge. As we have been trumpeting our sustainability bona fides through rankings, awards, and marketing, architects have off-loaded more and more responsibility for action to other sectors. There have been some legitimate improvements here and there, but meaningful change has not occurred, and, in absolute measures, the building sector has continued to increase energy consumption apace.
Strategies to reduce absolute energy demand have steadily disappeared from architectural teaching and practice over the decades. They have been replaced with the accounting sleight of hand that is “net-zero”—really a permission structure that allows us to increase the use of energy converting devices, to build larger and more luxuriously, and to continue our dependence on mechanical air conditioning systems. Many of my colleagues and acquaintances have “helpfully” asked if I knew that I could use solar panels and not have to worry about outages in the future. But there is no free energy. There is no energy use without a corresponding thermal emission, no matter the source. Every single energy exchange, including those that aren’t sourced by fossil fuels such as photovoltaics, releases heat as a byproduct. Greenhouse gases are problematic precisely because they are trapping these heat releases. The pace at which we are adding energy exchanges is far outstripping the pace at which we are decarbonizing. And even as we move toward cleaner and more efficient sources, we must use them to replace the dirtiest sources, rather than justify them to enable even more energy-consuming buildings and more energy-converting devices.
What is perhaps most alarming to me is the willful lack of recognition that water vapor is by far the largest greenhouse gas. While the increase in carbon dioxide through combustion has been the leading anthropogenic contributor to global warming, water vapor has not been addressed until recently. This could be because it is primarily due to natural processes and has a relatively short life in the atmosphere. But water vapor is now becoming an urgent issue due to the feedback cycle in which increasing temperature enables the atmosphere to hold more moisture, thereby increasing enthalpy and intensifying the greenhouse effect. Further compounding the natural process of evaporation is the water vapor emitted by cooling towers in electricity generation, as well as from air conditioning condensers. While a seemingly small contribution when compared to evaporation from oceans, these sources are a major contributor to urban heat islands, leading to an even greater need for mechanically assisted cooling. As the world continues to densely urbanize and rapidly expand the use of air conditioning, we can expect to see this spiral quickly toward greater uninhabitability. Carbon-free electricity generation might look good in the over-simplified calculations of net zero, but it is doing real and increasing harm if it is used to justify the expansion of cooling and other types of energy consumption. A radical reduction in energy demand must be our first line of action.
It is difficult for many to grasp how significant a role is played by water vapor but consider this example: 80 degrees Farenheit (27 degrees Celsius) air with 80 percent relative humidity contains more heat than 100 degrees Farenheit (38 degrees Celsius) air with 30 percent relative humidity.