Gary, Indiana Is a Warning: Our Power Grid Is More Fragile Than We Want to Admit

There is something deeply unsettling about an American city going without electricity for nearly two weeks.

Not a few hours.

Not an afternoon.

Not even a day or two.

Nearly two weeks.

That is what happened in Gary, Indiana, after a powerful derecho tore through Northwest Indiana on August 11, 2026. Winds approaching 100 miles per hour ripped through the region, knocking down trees, damaging infrastructure, and leaving hundreds of thousands of people without electricity. At the height of the disaster, roughly 370,000 customers were without power. While electricity was gradually restored across the region, Gary became one of the hardest-hit communities, with residents spending roughly two weeks dealing with the consequences of the outage.

And yes, the power is now largely back.

But that does not mean the problem is over.

In fact, I think the Gary outage should make us ask a much bigger question:

What happens if this becomes normal?

Because Gary is not simply a story about one storm, one utility company, or one city. It is a reminder of just how dependent modern civilization is on a system that most of us barely think about until it stops working.

We flip a switch and expect the lights to come on.

We plug something into an outlet and expect it to charge.

We open the refrigerator and expect the food inside to be cold.

We turn on the air conditioner when it is 90 degrees outside.

We expect traffic lights to work.

We expect gas stations to be able to pump fuel.

We expect grocery stores to be able to operate.

We expect hospitals to have electricity.

We expect cell towers and internet infrastructure to remain functional.

We expect water systems to continue pumping and treating water.

We expect banks, ATMs, businesses, schools, transportation systems and countless other pieces of modern society to continue operating.

All of those expectations ultimately depend on electricity.

And that is what makes a prolonged power outage so much more than an inconvenience.

It is a stress test for civilization.

Gary Shows What Happens When the Lights Stay Off

The August 11 storm was not some ordinary thunderstorm.

The derecho brought destructive winds, flooding and tornadoes across parts of Northwest Indiana. NIPSCO reported that more than 370,000 customers lost power. The damage was extensive enough that restoring electricity became a massive logistical operation requiring hundreds of lineworkers and vegetation crews.

But the thing that really caught my attention was the duration.

A power outage lasting several hours is annoying.

A power outage lasting a day is disruptive.

A power outage lasting several days becomes a serious emergency.

A power outage lasting nearly two weeks becomes something else entirely.

People lost refrigerated food. Businesses struggled to operate. Residents dealing with medical equipment requiring electricity faced an especially serious situation. Extreme summer heat made the lack of air conditioning more dangerous. Streets became dark. People had to seek assistance from shelters and community organizations.

Indiana officials reported nearly 99,000 outages statewide as of August 18, with the majority located in Northwest Indiana.

And this is where the science and infrastructure issue becomes important.

Because the question isn't simply whether a storm can knock out electricity.

Of course it can.

The question is how quickly can we recover when it does?

That is a much more important measurement of resilience.

A resilient electrical grid isn't one that never experiences failures. That is unrealistic. Trees fall. Lightning strikes. Tornadoes happen. Derechos happen. Hurricanes happen. Ice storms happen. Equipment fails.

A resilient grid is one that can absorb those shocks, isolate damaged sections, reroute electricity where possible, repair infrastructure quickly, and keep essential services functioning while restoration occurs.

Gary demonstrated what happens when that recovery process becomes painfully slow.

And that should concern all of us.

The Grid Is Not One Giant Machine

One misconception people sometimes have about the electrical grid is that it is basically one giant machine.

It isn't.

The American electrical system is an enormous interconnected network consisting of power plants, transmission lines, substations, distribution lines, transformers, switches, control systems and countless other components.

The system has different layers.

There is generation, where electricity is produced.

There is transmission, where large amounts of electricity travel long distances at high voltages.

Then there is distribution, which ultimately delivers electricity to homes and businesses.

A failure at one point doesn't necessarily mean the entire country goes dark.

That is actually one of the strengths of the system.

But it also means that there are countless potential points of failure.

A tree can take down a distribution line.

A tornado can destroy multiple poles.

Flooding can damage electrical equipment.

A substation can fail.

A transformer can malfunction.

Extreme heat can increase electricity demand.

Wildfires can threaten transmission infrastructure.

A cyberattack could potentially disrupt operations.

And increasingly severe weather can damage multiple pieces of infrastructure simultaneously.

That last part is particularly important.

The grid doesn't necessarily have to collapse because of one catastrophic failure.

Sometimes the problem is that many things fail at once.

That's what makes extreme weather so dangerous.

The Climate Connection Matters

There is another uncomfortable reality we have to discuss.

Extreme weather is part of the environment in which our infrastructure has to operate.

The atmosphere is changing as the planet warms, and that changes the conditions infrastructure has to withstand. That doesn't mean every individual storm can simply be blamed on climate change. Science doesn't work that way.

But it does mean we need to consider whether infrastructure designed around historical weather patterns is adequate for the conditions we are increasingly experiencing.

Gary's disaster is a good example of why.

The August 11 event involved extremely powerful winds, flooding and severe storms. The result was widespread infrastructure damage and a prolonged recovery.

And we have seen versions of this story repeatedly around the country.

Hurricanes can knock out electricity across entire regions.

Ice storms can bring down transmission and distribution lines.

Wildfires can threaten major infrastructure.

Heat waves can push electricity demand toward dangerous levels.

Winter storms can simultaneously increase electricity demand while disrupting generation and fuel supplies.

The problem isn't simply that extreme weather exists.

The problem is that our infrastructure is being asked to withstand extreme weather while simultaneously supporting an increasingly electrified society.

That is a difficult engineering problem.

And Electricity Demand Is Going Up

Here's another part of the problem that doesn't get nearly enough attention.

We are increasingly electrifying everything.

Electric vehicles.

Heat pumps.

Data centers.

Artificial intelligence.

Industrial processes.

Buildings.

Home appliances.

Everything from transportation to computing is becoming more dependent on electricity.

And that means the grid isn't simply being asked to maintain its current workload.

It is being asked to handle more.

The North American Electric Reliability Corporation, or NERC, has warned that electricity demand is expected to grow significantly over the next decade. Its 2025 Long-Term Reliability Assessment projects 224 gigawatts of summer peak demand growth over ten years, with data centers and the broader digital economy accounting for much of the increase. NERC also identifies growing resource adequacy concerns as demand increases faster than new resources and transmission can be added.

Think about what that means.

We are simultaneously saying:

"Let's electrify more of society."

"Let's build enormous data centers."

"Let's use more artificial intelligence."

"Let's transition transportation toward electricity."

"Let's rely on electric heating."

"Let's add more devices and technologies that require power."

And then we have to make sure the electrical system can actually support all of it.

That requires enormous investment.

It requires generation.

It requires transmission.

It requires distribution infrastructure.

It requires transformers.

It requires maintenance.

It requires trained workers.

It requires storage.

It requires backup systems.

It requires better forecasting.

And perhaps most importantly, it requires resilience.

Because having enough electricity under normal circumstances isn't the same thing as having a system that can survive extraordinary circumstances.

Reliability Is Not the Same as Resilience

This distinction is incredibly important.

Reliability generally asks:

Can we provide electricity when people need it?

Resilience asks:

What happens when something goes horribly wrong?

Those are related, but they aren't identical.

A system can work perfectly 99.9 percent of the time and still be disastrously vulnerable to a particular kind of event.

Imagine a bridge that works flawlessly every day but cannot survive a major flood.

It is reliable under normal circumstances.

It isn't particularly resilient.

The same principle applies to the electrical grid.

Gary's residents didn't care that the grid normally works.

They needed electricity after the storm.

They needed refrigerators.

They needed air conditioning.

They needed medical equipment.

They needed businesses to operate.

They needed traffic infrastructure.

They needed communication systems.

They needed their homes to function.

And when electricity stayed off for days and days, all of those systems became increasingly difficult to maintain.

That is resilience.

The Power Grid Is Becoming a Bigger Single Point of Failure

This is perhaps the biggest lesson I take from the Gary outage.

The more technologically advanced we become, the more dependent we become on electricity.

That sounds contradictory.

Technology is supposed to make society more resilient, right?

Sometimes it does.

But technology can also create additional dependencies.

Consider something as simple as paying for gasoline.

The fuel may physically exist in an underground storage tank.

But the gas station still needs electricity to pump it.

The payment terminal needs electricity.

The communications systems may need electricity.

Now consider grocery stores.

The food exists.

But refrigeration requires electricity.

Checkout systems require electricity.

Lighting requires electricity.

Computer networks require electricity.

Distribution centers require electricity.

Then consider water.

Water doesn't magically appear at your faucet.

Municipal water systems depend on pumps, treatment facilities, monitoring equipment and other infrastructure that require energy.

Then consider hospitals.

Modern medicine is incredibly dependent on reliable electricity.

Ventilators.

Monitors.

Imaging equipment.

Surgical equipment.

Refrigeration for certain medications.

Communications.

Computers.

Heating and cooling.

Backup generators can help, but generators aren't magic. They require fuel, maintenance and functioning infrastructure.

And suddenly you realize something:

An electrical outage isn't just an electrical problem.

It can become a food problem.

A water problem.

A transportation problem.

A communications problem.

A medical problem.

An economic problem.

A public safety problem.

A housing problem.

That's why the electrical grid deserves to be treated as critical infrastructure.

Because it is.

Gary Also Raises Questions About Inequality

There is another dimension to this story that I don't think we should ignore.

Not everyone experiences a prolonged power outage in the same way.

A wealthy household may have savings, a generator, backup batteries, multiple vehicles, a second property, the ability to stay at a hotel and the flexibility to work remotely.

A low-income household may have none of those things.

Someone living paycheck to paycheck can't necessarily replace hundreds of dollars of spoiled food.

Someone with mobility limitations may not be able to simply relocate.

Someone who depends on electrically powered medical equipment may face risks that another person doesn't.

Someone working hourly shifts may not be able to simply stop working because their neighborhood is without electricity.

Gary is a predominantly Black community that has also experienced decades of economic disinvestment and poverty. Reporting on the outage highlighted how those underlying conditions made the prolonged disaster particularly difficult for residents.

That means grid resilience isn't merely an engineering problem.

It is also a social problem.

Who gets restored first?

Which communities receive investment?

Which neighborhoods have the oldest infrastructure?

Who has the resources to prepare for an outage?

Who can afford backup power?

Who can afford to throw away spoiled food?

Who has somewhere else to go?

Those questions matter.

This Doesn't Mean the Grid Is About to Collapse

I want to be clear about something.

I'm not saying that Gary proves that the entire American electrical grid is about to collapse.

That would be sensationalism.

The United States has an enormous and highly sophisticated electrical system. Utilities and grid operators spend enormous amounts of time planning for failures and maintaining reliability.

And NERC's assessments don't say that the entire grid is doomed.

But they do identify growing risks.

NERC's 2025 assessment specifically warns about increasing resource adequacy concerns over the coming decade, including the challenges created by growing electricity demand and uncertainty about how quickly new resources can be built.

That's the more reasonable warning.

Not:

"The grid is going to collapse tomorrow."

But:

"The system is facing increasing stresses, and we need to invest before those stresses become crises."

Those are very different statements.

And the second one is supported by reality.

Gary Should Be a Stress Test

Instead of treating the Gary outage as something that happened and is now over, we should treat it as a giant real-world stress test.

What worked?

What failed?

How quickly were crews deployed?

Where were the bottlenecks?

How well did utilities communicate with residents?

How effective were emergency shelters?

How quickly could vulnerable residents receive assistance?

How much food was lost?

How did businesses cope?

Were there sufficient backup power systems?

Were emergency generators available where they were needed?

How well did local, state and federal agencies coordinate?

How much vegetation management had been performed before the storm?

What infrastructure failed?

And perhaps most importantly:

What can be changed before the next storm?

Indiana Gov. Mike Braun has already called for an investigation into NIPSCO's preparation and response to the outage.

That's appropriate.

Not because someone needs to be scapegoated.

But because disasters should teach us something.

If we experience a catastrophe and simply rebuild everything exactly as it was before, we haven't really learned anything.

We Need to Build for the World We Actually Have

This is the bigger lesson.

Infrastructure lasts decades.

The environment it operates in can change much faster.

A power line installed decades ago may still be perfectly functional.

A substation might still work.

A transformer might still operate.

A distribution network might technically meet its original design requirements.

But the question isn't whether the infrastructure worked when it was built.

The question is whether it can handle the conditions of today—and tomorrow.

That's why resilience has to become a core part of infrastructure planning.

We should be asking whether power lines should be buried in certain locations.

Whether vegetation management is sufficient.

Whether substations need additional protection.

Whether communities need more microgrids.

Whether hospitals, shelters and emergency facilities have sufficient backup power.

Whether battery storage can provide local resilience.

Whether distributed solar can help maintain critical services.

Whether utilities have enough trained workers to respond to large disasters.

Whether mutual assistance agreements can rapidly bring crews into devastated regions.

Whether communication systems can remain functional during prolonged outages.

And whether the grid is being expanded quickly enough to meet future demand.

These aren't glamorous questions.

Nobody gets excited about replacing a transformer.

Nobody posts viral videos celebrating upgraded distribution lines.

But this is the boring infrastructure work that keeps civilization functioning.

The Future Isn't Going to Be Less Electrified

If anything, the opposite is true.

The future is going to require more electricity.

More computing.

More electric vehicles.

More battery storage.

More electric heating.

More data centers.

More automation.

More communications.

More connected devices.

More artificial intelligence.

More electrically powered infrastructure.

And that's fine.

I actually think electrification can be an incredibly important part of building a cleaner and more technologically advanced society.

But we cannot electrify everything while treating the electrical grid like an afterthought.

That would be like building a massive highway system and refusing to maintain the bridges.

Eventually, something breaks.

And when it does, everyone discovers just how important the infrastructure really was.

Gary Is the Warning

That's ultimately what I think Gary, Indiana represents.

Not the end of the electrical grid.

Not proof that civilization is collapsing.

Not some inevitable prediction that America is going to spend weeks in darkness.

It is a warning.

A warning that extreme weather can overwhelm infrastructure.

A warning that restoration can take much longer than people expect.

A warning that vulnerable communities can suffer disproportionately.

A warning that our dependence on electricity is enormous.

A warning that electricity demand is growing.

And a warning that resilience cannot be something we think about only after the storm arrives.

The lights eventually came back on in Gary.

That is good.

People can begin rebuilding their routines.

Businesses can reopen.

Refrigerators can run again.

Air conditioners can turn back on.

Streetlights can illuminate neighborhoods.

Phones can charge.

Life can start returning to normal.

But we shouldn't let "normal" mean forgetting what happened.

Because the next Gary might not be Gary.

It could be another Midwestern city hit by a derecho.

It could be a coastal community hit by a hurricane.

It could be a Western community threatened by wildfire.

It could be a Southern city overwhelmed by extreme heat.

It could be a Northern community hit by an ice storm.

It could be somewhere none of us are thinking about right now.

And the question isn't whether the next major storm will knock out electricity somewhere.

It almost certainly will.

The question is how prepared we are when it happens.

Because our civilization runs on electricity.

And Gary, Indiana just gave us another reminder of what happens when the electricity stops.

**The lights coming back on should not be the end of the conversation.

It should be the beginning.