CBS Newsletter
Fall 1996
pg. 5

Urban Heat Catastrophes: The Summer 1995 Chicago Heat Wave

Figure 1: The bars indicate the numbers of deaths by race each day of the July 1995 heat wave in Chicago. The curve shows the heat index, which reflects the combined effect of temperature and humidity.

Last year's Chicago heat wave created a great deal of human discomfort and, according to the latest estimate, increased deaths in Cook County by more than 700 over five days. Epidemiological studies of heat-wave deaths have uncovered a number of socioeconomic, cultural, institutional, and physiological factors, but the role of the building and its interior conditions has been left largely unexamined.

A number of studies have compared ambient climate conditions to mortality during a heat storm. They found that there exists a heat-index threshold above which deaths increase rapidly. The duration, high humidity, high minimum temperatures, and low wind speeds all contribute to increased mortality, and a time lag exists between the peaks in the heat index and deaths. Surveys have also found that more deaths occurred in inner-city areas and disproportionately among older, infirm residents on the top floors of apartments without air conditioning. The mortality pattern appears to correlate with the thermal response of different building types to a heat storm, as well as to current conditions in the U.S. housing stock. Thus, it is surprising that studies have not been done on indoor conditions during a heat storm or on what role the thermal characteristics and operations of buildings play in this increased mortality.

The heat storm that affected Chicago from July 12 to 16, 1995, was a particularly acute episode containing all the danger signals-high temperatures and humidity and low wind speeds over a five-day period. The heat index reached 118°F (48°C) on the hottest day.

The Physiology of Human Discomfort and Heat Stress

Human discomfort and heat stress depend on temperature, humidity, radiant heat gain, and wind. A commonly used index for determining heat illness risk is the WBGT, a heat index that combines wet bulb, black globe, and dry globe temperatures into a single measure of environmental heat (Figure 1). Below a heat index of 64°F (18°C), the risk of heat injury is small. Above 82°F (28°C), strenuous activity should be avoided. From 90 to 104°F (32 to 40°C), heat cramps and exhaustion are possible; from 105 to 130°F (41 to 54°C), cramps and exhaustion are likely and heat stroke is possible; and above 130°F, heat stroke is highly likely. During the five days of the July heat storm in Chicago, 60 percent of the hours were above a heat index of 90°F, and 20 percent above 104°F.

The impact of excessive temperature on mortality is difficult to quantify and varies with the physical condition of the person. Preliminary statistics from the Chicago heat storm indicate that most of the victims were elderly people who were already in a weakened state.

Computer Simulations of Interior Conditions

The DOE-2 building simulation program was used to simulate indoor conditions in four prototypical multifamily buildings of different vintages during the July 1995 heat storm in the absence of air conditioning. The buildings were simulated first with windows closed, and then with windows opened for ventilation whenever the temperatures outdoor were lower than inside. To study the benefits of potential conservation strategies, the simulations were repeated with additional ceiling insulation, light-colored roofs, and lowered window-shading coefficients. These cases are called the "weatherized buildings."

If the buildings were unventilated, as was often reported to be the case, indoor temperatures could reach 108°F (42°C) on the top floors of buildings built in the 1940s (Figure 2). They were hotter than human body temperature for 80 percent of the hours during the peak three days. Conditions in the 1970s apartment building were even worse, with temperatures averaging 108°F over the three-day period. Because of their greater mass and moderate insulation, these buildings would remain hot for days after the peak air temperatures had passed. The heat index reached 129°F (54°C) in the 1940s apartments and 134°F (57°C) in the 1970s apartments over the period.

Strategies to Prevent Excessive Overheating during a Heat Storm

The simulations show that the single most important strategy to prevent excessive building overheating during a heat storm is ventilation. Under such conditions, ventilation will not keep the buildings comfortable, but will prevent them from acting like solar ovens and keep temperatures indoors close to or below those outdoors. In older, uninsulated buildings, adding ceiling insulation and lightening the roof color will have an appreciable impact on conditions in top-floor apartments. However, such weatherization in newer buildings will have a minimal impact on their indoor conditions during a heat storm.


The prevention or reduction of mortality during a heat storm should be viewed as a form of disaster relief. Because of the public outcry over the 1995 heat storm, the city of Chicago has been working on a relief plan. So far, their efforts have focused on public outreach, providing warnings, checking on tenants, and moving people to cool rooms to escape the heat. This preliminary study suggests the dangers can also be lessened by improving the thermal conditions and operations of the buildings. Since heat storms are sporadic, any such weatherization effort must consist of simple and inexpensive strategies.

Figure 2: Computer-simulated indoor temperatures in the top floor of a prototypical 1940s two-story apartment building in Chicago during the July 1995 heat wave. In the existing building, top-floor temperatures reached 108°F and remained high even after the outdoor temperatures had started to drop. The addition of attic insulation, white paint on the roof, and open windows brought top-floor temperatures in line with outdoor temperatures.

—Joe Huang

Info icon

Joe Huang
Energy Analysis Program
(510) 486-7082; (510) 486-6996 fax

EETD Newsletter Home Page
CBS Newsletter Home Page
Table of Contents for this Issue