How the Lights Stay On

Chapter 17: The Grid You’ll Build

If you are reading this book in a house with electricity, the grid is, right now, working for you.

A turbine is spinning somewhere. A wire is humming somewhere. A dispatcher is watching a number on a screen somewhere. A transformer is converting voltage somewhere. A control room is quietly running. The lights are on. You probably are not thinking about any of it.

If you are reading this book in 2030, much of what you just learned is still true. The grid is still mostly the same machine, still mostly running on the same physics, still mostly operated by the same kinds of people. Some new things have been built. Some old things have been retired. The basics have not changed.

If you are reading this book in 2040, more has changed than you might expect. The reactors at the Crane Clean Energy Center are humming. Several small modular reactors have been operating for a decade. Several have failed. The duck curve has been mostly tamed by storage in some states and not yet tamed in others. Loudoun County’s data centers are several times bigger than they are now. The XFRA pilots have either become routine or quietly disappeared.

If you are reading this book in 2050, the grid will be a thing your generation built.

This chapter is about that.

Let’s take a quick inventory of what we have learned.

The grid is one giant synchronized machine. Electricity is made and used in the same instant. The frequency is the heartbeat — 60 Hz — and it tells you whether supply and demand are in balance. Dispatchers in control rooms watch the heartbeat, every minute of every day.

Transmission lines move electricity from where it is made to where it is used. They have to be enormous because the physics of resistance demands high voltage for long distance. Building new ones takes too long because of permitting, not technology. The same is true of substations, distribution upgrades, and almost every other piece of physical grid infrastructure.

Renewable energy is now the cheapest electricity per kilowatt-hour in most of the world. But it is intermittent, and a grid cannot run on intermittency alone. Storage helps. Nuclear helps. Transmission helps. None of these is fully solved at the scale required.

The grid is becoming smarter. Sensors, smart meters, demand response, distributed energy resources, virtual power plants — all of these are real, partially deployed, and growing. Some homes are now power plants. Some are now data centers.

Different countries have made different bets on the trilemma. None of them has all three corners. Each has paid different prices. The conversation about which corner to sacrifice is ongoing in every state in the United States and every country in the world.

And underneath all of it, tens of thousands of people are doing the actual work. Most of them are graying. Most of their successors have not yet been trained.

That is the picture as of 2026. It will change.

There are, in 2026, a handful of unsolved problems that the next twenty years of energy work will mostly be about.

These are not minor. They are the problems that will determine whether the world can build a grid that is genuinely clean, genuinely affordable, and genuinely reliable — or whether some corner of the trilemma will continue to be sacrificed because we did not figure out how to do better.

I want to name them, because the people who solve them will probably be reading this book. Or books like it. Or being taught by people who read this book.

Long-duration storage. Lithium batteries handle four hours. Pumped hydro handles a day, but only in places with mountains. We do not have an economical way to store electricity for a week, let alone a season. Without long-duration storage, a grid running mostly on solar and wind cannot reliably handle a windless cloudy week in February. Iron-air batteries, hydrogen, compressed air, thermal storage, gravity storage — all of these are candidates. None has yet been deployed at the scale required. Whoever solves this gets to redesign the grid.

Transmission permitting. The wires we need to build between now and 2050 are wires that have not yet been planned, on land that has not yet been crossed, through political and legal fights that have not yet been won. The technology is not the bottleneck. The bottleneck is the seventeen-year permitting timeline of the SunZia project, the eight-year fight over Northern Pass, the lawsuits and counter-lawsuits and environmental reviews of every major line proposed in the last twenty years. Somebody has to figure out how to build new transmission corridors in years rather than decades — without trampling the communities, ecologies, and property rights that have legitimately resisted them in the past.

Grid-forming inverters. The synchronized machine from Chapter 2 has, for a hundred years, been held together by the inertia of massive spinning generators. Solar and battery inverters do not have inertia. Engineers are working on inverters that can mimic inertia and even stabilize a grid without any spinning mass at all. The technology is improving fast but not yet deployed at scale. A grid that can run on solar, wind, and batteries alone — without any rotating thermal plants — is not yet engineering reality. It will be.

Demand-side coordination at scale. The smart grid is partially deployed. Demand response programs exist. Virtual power plants exist. But the system that would coordinate billions of small assets — every thermostat, every EV charger, every water heater, every Powerwall — in real time, across the entire country, with appropriate privacy and security safeguards, has not been built. Most of the technology exists. The systems engineering does not.

Cybersecurity. A digital grid is a hackable grid. The Ukrainian attacks in 2015 and 2016 showed what is possible. The defensive work is ongoing, and the attackers are also getting better. This will not be solved once. It will be a permanent arms race for the rest of the century.

Workforce. Every problem above requires people to solve it. The workforce is graying faster than it is being replaced. Closing the gap is going to require apprenticeships, community colleges, university engineering programs, and a cultural shift in how Americans think about working on physical infrastructure. The pay is decent. The work is meaningful. The training programs exist or can be built. The bottleneck is whether enough young people choose to do this work.

Cost-effective firm clean power. The deepest open problem. Nuclear is firm and clean but expensive. Geothermal is firm and clean but limited to certain geologies, at least so far. Long-duration storage might solve it but is not yet proven at scale. Carbon capture for gas plants might solve it but is not yet economic. We need a source — or some combination of sources — that delivers reliable electricity 24 hours a day with low carbon emissions at competitive cost. We do not yet know which technology will win. Whoever figures it out will, in the most literal sense, build the grid of the 2050s.

If you are reading this book in your teens, you are looking at a generational opportunity.

The grid your generation will inherit is being designed, right now, by people who will mostly be retired by the time you are running it. The choices being made today — to build this gas plant or that nuclear reactor, to permit this transmission line or block it, to invest in batteries or hydrogen, to upgrade the smart grid or hold back — will be your inheritance.

But the choices being made next decade will be the ones you make.

You do not have to become an electrical engineer to be part of this. The grid needs lawyers who specialize in permitting reform. It needs policy experts who can navigate state utility commissions. It needs entrepreneurs who can build the companies that will deploy the technology. It needs journalists who can explain to the public what is actually happening. It needs financial people who can structure the deals that fund the infrastructure. It needs urban planners who can think about land use in the era of distributed energy. It needs writers who can tell the stories that change minds.

It also, very much, needs the engineers — the chemists working on iron-air batteries in Massachusetts, the nuclear engineers building Hermes reactors in Tennessee, the software architects writing the next generation of grid management systems, the meteorologists training AI to forecast wind speeds eighteen hours out. Those jobs are growing faster than they can be filled.

The grid your generation builds will be different from the one you grew up on. It will probably be cleaner, almost certainly more digital, possibly more distributed, definitely more contested. It will not be perfect. The trilemma will still apply.

But it will, in significant ways, be yours.

Let me close where this book started.

In Chapter 1, I told you about the Blink — the small flicker of my Hoboken apartment’s bulbs every time the air conditioner turned on. I told you that the Blink contained, in miniature, the same physics as the 2003 Northeast blackout. The same balance act. The same juggling of supply and demand, in a fraction of a second.

You have, since then, met the synchronized continent-sized machine that produced the Blink. You have walked through transmission lines and substations and pole transformers. You have read about the duck curve and the lithium batteries that are starting to flatten it. You have heard the story of the FirstEnergy operators in 2003 and the ERCOT operators in 2021. You have walked along the Susquehanna River and watched the cooling towers prepare to come back to life. You have stood in Loudoun County and looked at the office park where so much of the modern world’s data lives. You have learned the trilemma, and you have learned to apply it to any energy claim you ever encounter.

You have also met the people. The dispatchers. The lineworkers. The plant operators. The transmission planners. The forecasters. The cybersecurity teams. The tree-trimming crews. The office boy who drove everyone home on August 14, 2003.

So now, when you flip the switch and the lights come on, you know what is happening on the other side. A turbine somewhere is spinning. Wires somewhere are humming. A dispatcher somewhere is watching a number. A transformer somewhere is doing its small piece of the work. Tens of thousands of people, most of whom you will never meet, are keeping the grid alive for one more minute.

You also know that this is not stable. The grid is being rebuilt. The technologies are changing. The fuels are changing. The financial structures are changing. The workforce is turning over. The trilemma is being re-negotiated.

Pay attention next time you flip a switch.

The whole world is on the other side of it.

Some of it works because of choices made a hundred years ago. The rest is being decided now, by people not much older than you are.

And the rest of the rest will be decided by you.

All 18 chapters of How the Lights Stay On · Library · All courses