Denise Fair Razo runs Detroit's Health Department with 250 staff and a $50 million budget. She led the city's COVID response, serves on Michigan's Public Health Advisory Council, manages everything from vaccination clinics to emergency preparedness. By any reasonable measure, she's good at her job.

Last June, when five days of mid-to-upper-90s heat settled over Detroit, her department did what competent public health departments do. They activated the cooling centers. Extended hours at three recreation facilities. Public libraries during normal operating hours. Press releases about staying safe and checking on neighbors.

Twelve locations total. For 630,000 people. None of them have backup power.

Running a heat emergency response designed for a climate that no longer exists means you can do everything right by the old standards and still be catastrophically unprepared for what's coming. Fair Razo isn't failing at her job. The job itself has become obsolete faster than the institution can recognize.

A 2023 study published in Environmental Science & Technology modeled what happens when extreme heat coincides with widespread power failure in Detroit—the kind of compound event that's becoming more likely as aging Midwestern infrastructure buckles under climate pressures it was never designed to handle. The findings: between 68 and 100 percent of Detroit's population would face elevated risk of heat exhaustion or heat stroke. Mortality rates could more than double. 450,000 people endangered in a scenario researchers compared to Hurricane Katrina.

Twelve cooling centers can maybe handle 6,000 to 12,000 people if you pack them in and rotate folks through. One to two percent of the population. Which is fine if you're managing heat waves the way Detroit thought about them in 1995 or 2005—occasional inconveniences where vulnerable people need a place to cool off for an afternoon.

It's not fine if you're responsible for 450,000 people when the heat and the blackout arrive together and none of your cooling centers can operate without electricity.

When the Grid Goes Down

Michigan ranks fourth nationally for annual customer outage hours—8.5 hours compared to a national average of 4.5. In 2023, almost 45 percent of DTE Energy customers suffered eight or more hours of service disruptions during summer heat waves. The same heat that makes cooling centers necessary is what makes the grid most likely to fail.

Everybody cranks their AC at once, the overhead wires heat up and sag until they hit a tree branch, and suddenly 45,000 customers are sitting in the dark while the temperature climbs. DTE reports a 70 percent improvement in reliability between 2023 and 2024, thanks to $1.5 billion in grid investment and—this is the telling part—"less extreme weather."

Less extreme weather isn't an infrastructure strategy. It's luck.

The compound event scenario assumes three consecutive days of heat index values exceeding 103°F. Exactly what climate projections show becoming more frequent across the upper Midwest. Not someday. Now. The June 2024 heat wave that prompted Fair Razo's cooling center activation was potentially the worst Detroit had seen in twenty years.

The researchers used building energy models to simulate what happens to interior temperatures when the power goes out. In Detroit, the greatest heat exposures would be in one-story, single-family houses—the housing stock that dominates much of the city. Average indoor temperatures during the modeled event: 92°F.

Your housing type determines your survival odds.

86 percent of Detroit neighborhoods experience an increase in temperature of up to 8°F or more from the urban heat island effect. That effect is most pronounced at night, exactly when the body needs to cool down. The neighborhoods with the worst heat island effect, the oldest housing stock, the least air conditioning, the most vulnerable residents—they're the same neighborhoods that have been systematically divested from for decades. The heat wave doesn't create the inequality. It just makes visible what was already there.

In 1995, Chicago's heat wave killed over 700 people. The city had five cooling centers and knew it had a problem. Most of the dead were elderly poor who either had no air conditioning or couldn't afford to run it. Detroit has twelve centers for a population six times larger and still thinks it's a cold-weather city that occasionally gets hot.

Doing Everything Right

During that June heat wave, Fair Razo's cooling centers operated weekdays from 8 a.m. to 8 p.m. Three facilities with extended weekend hours. A reasonable response to a heat wave as Detroit has historically understood heat waves. She issued the right statements:

True enough. Inadequate anyway.

The researchers found that 50,000 Detroit residents face elevated risk during heat waves even without a blackout, due to lack of air conditioning, housing quality, and underlying health conditions. Add widespread power failure lasting three days. The cooling centers can't operate. The people who need them most—elderly, poor, living in the heart of the city—are the ones least likely to reach them during a grid emergency.

None of the three cities in the compound event study—Detroit, Phoenix, Atlanta—has developed an emergency response plan to manage widespread power outages under heat wave conditions. The researchers concluded that such events would require "a far more extensive network of emergency cooling centers than is presently established in each city, with mandated back-up power generation."

What you have now isn't remotely adequate for what's coming.

Fair Razo did everything right by the standards of 2015. Competence at managing the old emergencies doesn't prepare you for the new ones. She's running a heat wave response designed for a climate that no longer exists, and when the compound event arrives, her twelve cooling centers without backup power will be about as useful as the Titanic's lifeboats. Adequate for the disaster they imagined, catastrophically insufficient for the one they got.

Cities that thought they were climate refuges are becoming climate traps. The same industrial infrastructure that built Detroit—the density, the housing stock, the grid designed for different weather—now makes it vulnerable in ways Sun Belt cities at least recognize and plan for, however inadequately.

The modeling keeps showing what happens when the heat wave and the blackout arrive together. A competent public health officer keeps doing a job that's becoming obsolete faster than the institution can adapt. She knows the math. Twelve cooling centers. 630,000 people. 450,000 at risk.

She's just running a system that hasn't yet admitted what it's becoming.