why is the power grid failing

You already know the grid is unreliable. You've sat in the dark enough times to know that. But "the grid is bad" is a feeling, not an argument. What follows is the argument. Numbers. Sources. Data you can check yourself. No politics. No conspiracies. Just the documented, measurable, boring-on-a-spreadsheet reasons why the American power grid is failing more often and why every trend line says it's going to get worse.

I spent months pulling data from the Department of Energy, NERC filings, utility commission reports, ASCE infrastructure assessments, and about a dozen peer-reviewed studies. What I found wasn't surprising. But seeing it all in one place was genuinely unsettling.

the numbers: outages are up 64% and accelerating

Let's start with the headline number. According to the Department of Energy's data on major power disruptions (tracked through form OE-417), the frequency of significant outage events has increased approximately 64% since the year 2000. That's not my estimate. That's the government's own data.

But the overall number actually understates the problem, because the trend isn't linear. It's accelerating. From 2000 to 2013, outages increased at a moderate pace. Since 2013, the curve has steepened dramatically. The five-year rolling average of major outage events from 2019 to 2024 is roughly double what it was from 2005 to 2010.

And these are just the big ones. DOE form OE-417 only captures events affecting 50,000 or more customers or involving significant disruptions to the bulk power system. The small outages, the ones where your neighborhood goes dark for six hours because a transformer blew or a tree fell on a line, aren't in these numbers. The Institute of Electrical and Electronics Engineers (IEEE) tracks those separately and the picture is even worse. The average US electricity customer experienced about 8 hours of power interruption in 2022, compared to roughly 2 hours for customers in Germany or Japan.

Eight hours per customer per year. In the richest country in the history of the world. That's four times worse than most developed nations.

the infrastructure is older than it was ever supposed to get

The backbone of the American power grid was built during the great infrastructure buildout of the 1950s, 1960s, and 1970s. The engineers who designed it were good at their jobs. They built things to last 30 to 40 years, which was the standard design life for large power transformers, transmission lines, and substation equipment.

Those 30 to 40 years have come and gone. The equipment is still there.

The average large power transformer in the United States is now over 40 years old. These are the critical pieces of equipment that step voltage up and down between generation, transmission, and distribution. There are roughly 80,000 of them in the US. The majority have exceeded their intended design life. They don't stop working the day they turn 40. They degrade. Insulation breaks down. Oil quality deteriorates. Failure rates increase exponentially past design life. And when a large power transformer fails, it doesn't just blow a fuse. It can cause fires, explosions, and cascading outages. Replacement takes 12 to 18 months because these things are custom-manufactured, often overseas.

Over 70% of transmission lines in the US are at least 25 years old. Many are approaching 50. The towers, conductors, insulators, and hardware all degrade with age, weather exposure, and electrical stress. The American Society of Civil Engineers gave the nation's energy infrastructure a C- in their most recent report card. A C-. For the system that everything else depends on.

The distribution system, the poles and wires that actually deliver power to your house, is in even worse shape. Wooden utility poles have a design life of 30 to 40 years. There are approximately 180 million of them in the United States. The replacement rate is not keeping pace with degradation. Many utilities are running on a reactive maintenance model: they replace things when they fail, not before. That's not a maintenance strategy. That's a prayer.

demand is surging and the grid wasn't built for it

For about two decades, from roughly 2000 to 2020, US electricity demand was essentially flat. Energy efficiency improvements in appliances, lighting (the LED revolution), and building codes roughly offset population growth and economic expansion. Grid planners got comfortable with the idea that demand wasn't really going up.

That era is over.

Three forces are now pushing demand growth at rates the grid hasn't seen since the mid-20th century:

Electric vehicles. The US had about 4 million EVs on the road at the end of 2024. The projected number by 2030 ranges from 26 to 50 million depending on which forecast you use. Each EV adds roughly 3,000 to 4,000 kWh of annual electricity demand, equivalent to adding a small house to the grid. The charging infrastructure, especially DC fast chargers, creates concentrated load spikes that the local distribution system often can't handle without upgrades.

Data centers. This is the big one that crept up on everyone. The explosion of cloud computing, and now AI training and inference, has turned data centers into some of the largest electricity consumers in the country. The Electric Power Research Institute estimated that data center electricity consumption could double or even triple by 2030 compared to 2023 levels. Individual data center campuses now draw hundreds of megawatts, equivalent to small cities. And they need that power 24/7/365 with essentially zero tolerance for interruption.

In northern Virginia, the largest data center market in the world, Dominion Energy has warned that demand is growing faster than they can build generation and transmission capacity. The same story is playing out in Georgia, Texas, Ohio, and anywhere else the hyperscalers are building. Grid operators in multiple regions have flagged data center demand as a reliability risk.

Electrification of everything else. Heat pumps replacing gas furnaces. Induction cooktops replacing gas ranges. Industrial processes shifting from fossil fuels to electricity. Each of these is individually modest, but collectively they represent a significant and sustained increase in electricity demand on a grid that was sized for a different era.

The North American Electric Reliability Corporation (NERC) projected in its 2024 Long-Term Reliability Assessment that peak electricity demand could increase by 38% over the next decade in some regions. Let that number sink in. Nearly 40% more demand on a grid where the average transformer is already past its design life.

extreme weather is breaking things faster than we can fix them

Weather-related outages are the single largest cause of major power disruptions in the United States. That's been true for at least two decades, and the trend is getting worse. DOE data shows weather-related outage events have roughly doubled since the early 2000s.

This isn't about one type of weather. It's everything, simultaneously getting more intense:

Hurricanes are intensifying. The proportion of hurricanes reaching Category 4 or 5 intensity has increased. Rapid intensification events, where a storm strengthens dramatically in a short period, are becoming more common. Hurricane Helene in 2024 demonstrated that the destruction extends hundreds of miles inland. The history of hurricane-driven outages is long and it keeps getting new entries.

Ice storms are still devastating. The 1998 North American Ice Storm collapsed over 1,000 transmission towers. Ice loading on power lines remains one of the most destructive weather phenomena for grid infrastructure, and the frequency of significant icing events has not decreased.

Heat waves are killing the grid. Extreme heat reduces the efficiency of power generation (thermal plants produce less power when cooling water is warm), increases transmission losses (lines sag and lose capacity in heat), and spikes demand (air conditioning). The grid faces its highest stress during heat waves, which is exactly when it's least capable of performing. The deadly 2023 heat dome across the southern US pushed multiple grid operators to their limits.

Wildfires create a vicious cycle. In California and increasingly across the West, utilities are now deliberately shutting off power during high fire-risk conditions to prevent their equipment from starting wildfires. Pacific Gas & Electric's Public Safety Power Shutoffs have affected millions of customers. So the choice becomes: risk your power line starting a wildfire that destroys a town, or deliberately cut power to hundreds of thousands of people. Neither option involves reliable electricity.

Winter storms hit systems designed for mild weather. Texas in 2021 proved that cold weather can be just as destructive as hurricanes when the grid isn't prepared for it. But Texas isn't alone. Winter storms across the Southeast, Mid-Atlantic, and Pacific Northwest regularly cause multi-day outages because distribution systems in those regions aren't hardened against ice and wind loading to the same degree as northern states.

The climate that the grid was engineered for, the historical climate data that engineers used in the 1960s and 1970s when they designed this infrastructure, no longer describes the climate the grid operates in. The design parameters are wrong. Not slightly wrong. Meaningfully wrong. And updating the physical infrastructure to match current and projected climate conditions would require the kind of investment that nobody seems willing to make.

utilities have no financial incentive to fix this

This is the part that makes me the angriest, and it's the part that gets the least attention.

Most electric utilities in the US operate as regulated monopolies. They're guaranteed a rate of return on their capital investments, typically 9% to 11%, approved by state public utility commissions. This is called the "rate base" model, and it creates a perverse incentive structure that actively works against grid reliability.

Here's how it works: utilities earn their profit as a percentage of their invested capital. They make more money by building new things than by maintaining old things. A utility that spends $100 million on a new substation earns $10 million per year on that investment. A utility that spends $100 million on maintaining existing equipment earns the same return, but the regulatory process for approving maintenance spending is often more contentious and harder to pass through than new construction.

The result is predictable. Utilities build new things when politically expedient and defer maintenance on existing infrastructure until it fails. The maintenance backlog grows. The failure rate increases. Customers experience more outages. And when things break catastrophically, the utility files for an emergency rate increase to fund repairs, and customers pay for it.

You pay for the infrastructure. You pay for the inadequate maintenance. You pay for the failure. And you pay for the repair. The utility profits at every stage.

When utilities are fined by NERC or state regulators for reliability violations, those fines are frequently passed through to ratepayers as well. So you're also paying for the penalty assessed against the company that failed to maintain the system you're paying them to maintain. It's a closed loop of accountability that never actually reaches the people making decisions.

Executive compensation at the largest US utilities has roughly doubled over the past 15 years. Reliability metrics, during that same period, have declined. There is no meaningful financial consequence for utility executives when the grid fails. There are enormous financial rewards for keeping costs down and shareholder returns up.

NERC keeps warning. nobody keeps listening.

The North American Electric Reliability Corporation exists specifically to monitor grid reliability and warn about emerging risks. They publish detailed assessments every year. Reading them is like reading a series of increasingly urgent memos from someone who knows the building is on fire and can't get anyone to evacuate.

NERC's 2024 Long-Term Reliability Assessment identified several "high risk" areas:

• Insufficient generation reserves. Multiple regions are projected to have inadequate generation capacity to meet peak demand within the next five years. Retirements of coal and aging natural gas plants are outpacing the construction of replacement generation.
• Transmission inadequacy. The transmission system, the high-voltage backbone that moves power from where it's generated to where it's consumed, is not being expanded fast enough to accommodate new generation sources (particularly renewables, which are often located far from population centers) or growing demand.
• Inverter-based resource integration. The rapid addition of solar, wind, and battery storage changes the electrical characteristics of the grid in ways that existing protection systems and operating procedures weren't designed for. Grid stability depends on inertia, and inverter-based resources provide less of it than traditional spinning generators.
• Extreme weather vulnerability. NERC explicitly warned that the grid's vulnerability to extreme weather events is increasing and that current planning standards do not adequately account for the frequency and intensity of events now being observed.

These aren't fringe warnings from outsiders. This is the official reliability organization of the North American power grid saying, in writing, that the system is heading toward more failures. And the response from utilities and regulators has been, broadly speaking, to acknowledge the warnings and continue operating largely as before.

the transmission bottleneck

Even if we had unlimited money and political will to build new generation capacity, there's a bottleneck that would stop us: the transmission system.

Building a new high-voltage transmission line in the United States currently takes 10 to 15 years from proposal to energization. That timeline includes permitting, environmental review, right-of-way acquisition, legal challenges, and construction. In some cases, it takes longer. Projects routinely spend more years in permitting than in construction.

The interconnection queue, the backlog of new generation projects waiting to connect to the grid, reached over 2,600 gigawatts by the end of 2024 according to Lawrence Berkeley National Laboratory. That's more than twice the current installed generation capacity of the entire US. The vast majority of those projects are solar, wind, and battery storage. They can't connect because there isn't enough transmission capacity, and building new transmission capacity takes a decade.

This bottleneck affects reliability directly. When generation capacity retires (coal plants closing, old gas plants shutting down), replacement capacity needs to connect to the grid through the transmission system. If the transmission system can't accommodate it, the region ends up with less available generation than it needs. That's how you get the "insufficient generation reserves" warnings from NERC.

It also means that even areas with abundant renewable energy potential can't get that power to the population centers that need it. West Texas has some of the best wind resources in the country, but the transmission lines to move that power to Dallas and Houston are often congested. Same story with solar in the desert Southwest and demand in Southern California. The generation exists. The wires to move it don't.

regional vulnerabilities: where the grid is weakest

The US doesn't have one grid. It has three: the Eastern Interconnection, the Western Interconnection, and ERCOT (Texas). Each has its own vulnerabilities.

ERCOT: isolated by design, vulnerable by consequence

Texas deliberately operates its own grid, largely separate from the rest of the country, specifically to avoid federal regulation. This means that when ERCOT runs short on power, it can't easily import electricity from neighboring states. During Winter Storm Uri in 2021, this isolation nearly caused a total grid collapse that would have taken weeks to recover from. Texas has made some winterization improvements since Uri, but the fundamental isolation remains, and so does the vulnerability. The state's explosive growth in data center demand is adding further strain.

the Northeast: old infrastructure, dense demand

The northeastern US has some of the oldest grid infrastructure in the country, combined with dense urban demand and limited space for new construction. ISO New England and NYISO (the grid operators for New England and New York) have repeatedly warned about winter reliability risks as natural gas plants that serve both heating and electricity compete for limited pipeline capacity. The region is retiring nuclear plants and struggling to build replacement capacity fast enough. Transmission constraints make it difficult to move power from areas with surplus to areas with deficit.

California: wildfires, shutoffs, and the duck curve

California faces a unique combination of reliability challenges. Wildfire risk forces utilities to conduct Public Safety Power Shutoffs that proactively cut power to millions. The state's aggressive renewable energy targets have created the "duck curve" problem, where solar generation floods the grid midday and then drops off sharply in the evening just as demand peaks, requiring massive ramp-up from other sources. The 2020 rolling blackouts during a heat wave demonstrated that the state's capacity margins are razor-thin. And the push to electrify everything, from cars to heating, is increasing demand on a system already under stress.

the Southeast: hurricane alley with aging distribution

The southeastern US faces annual hurricane threats combined with extensive overhead distribution infrastructure (wooden poles and wires) that is highly vulnerable to wind and falling trees. The region also has some of the fastest population growth in the country, adding demand to systems that need significant upgrades. Hurricane Helene in 2024 demonstrated that the threat extends deep inland, into mountain communities that had never planned for hurricane-scale damage.

the Midwest: extreme temperature swings

The Midwestern grid faces stress from both extreme cold (polar vortex events) and extreme heat. MISO, the grid operator for much of the Midwest, has warned that capacity margins are tightening as coal plants retire. The 2014 polar vortex brought the region to the brink of widespread blackouts, and similar events are projected to become more frequent.

so what's actually going to happen

I'm not a fortune teller. But the data supports some straightforward projections:

Outages will continue to increase in frequency and duration. Every major driver of outages, aging infrastructure, rising demand, extreme weather, is trending in the wrong direction. The investment needed to reverse these trends is measured in trillions of dollars and decades of construction. It is not happening at the required scale.

Some regions will get meaningfully worse before they get better. Areas with rapidly growing demand (Texas, the Southeast, data center corridors), areas with aging infrastructure and limited new construction (the Northeast), and areas with extreme weather exposure (everywhere, increasingly) will see the sharpest reliability declines.

Electricity prices will rise. Whether utilities invest or don't invest, prices are going up. If they invest, the capital cost is passed to ratepayers. If they don't invest, the repair cost from failures is passed to ratepayers. Either way, you're paying more for less reliable power.

The gap between what the grid provides and what people need will widen. As more of daily life depends on electricity (EVs, remote work, medical devices, home heating), the consequence of each outage becomes more severe. A four-hour outage in 1990 meant your TV went off. A four-hour outage in 2026 means your car can't charge, your work stops, your thermostat goes dead, and your medical equipment needs a backup plan.

None of this is speculative. It's the trajectory of documented trends, extended forward at their current rates. The grid will not fix itself. The utility companies will not fix it for you. The government is not going to rebuild the grid in time to matter for your family.

What you can do is stop depending entirely on a system that is demonstrably, measurably, objectively not dependable.

Start with the history of major outages if you want to see the pattern in full. If you're ready to act, the whole-home generator guide and the generator vs. battery backup comparison are where I'd go next. Or browse all the guides and find what fits your situation.