
How The Grid Works
Fuel becomes rotation, rotation becomes current, and current is pushed up to transmission voltage and back down again. Underneath all of it is one rule: production must equal consumption, continuously.
How electricity is made
Most electricity on the grid still comes from a spinning machine. Heat from nuclear fission, natural gas, coal or concentrated sunlight boils water into steam, or expands combustion gases directly, and the pressure turns a turbine. The turbine turns a generator, and a rotating magnetic field inside it induces alternating current in the surrounding windings. Hydroelectric plants and wind turbines skip the heat and use moving water or air to turn the shaft.
Solar panels and batteries work differently. They produce direct current, which an inverter converts into alternating current before it reaches the grid. The distinction matters more than it sounds. A spinning generator carries physical inertia: its mass resists sudden changes in speed and therefore in frequency, which buys operators a few seconds when something goes wrong. Inverters have no moving mass, so as their share of the system grows, that stabilizing response has to be created deliberately in software and by other equipment.
Synchronous
Spinning generators
Nuclear, gas, coal, hydro, geothermal and most large-scale plants. Locked in step with grid frequency; their rotating mass provides inertia and short-circuit strength.
Inverter-based
Solar, batteries, most modern wind
Power electronics synthesize the AC waveform. Fast and precise, but inertia and voltage support must be engineered in rather than coming for free.
Why voltage goes up before it comes down
Moving power through a wire wastes some of it as heat, and the waste rises with the square of the current. For a given amount of power, raising the voltage lowers the current proportionally, so a line at 500 kV loses a small fraction of what the same power would lose at distribution voltage. This is the entire reason transformers exist, and the reason a power plant’s output is stepped up the moment it leaves the generator.
Bar lengths are illustrative of relative voltage, not to scale. Roughly five percent of the electricity generated in the United States is lost in transmission and distribution before it reaches a customer.
Which plants run, and in what order
Operators do not run every plant at once. They stack available generators from cheapest to most expensive to operate and work up the stack until supply meets demand. That ordering is called merit order or economic dispatch, and the cost that matters is the marginal cost of producing one more unit, mostly fuel and variable maintenance, not the cost of having built the plant.
Resources with no fuel cost sit at the bottom and run whenever they are available. Nuclear sits near the bottom as well and runs continuously, because its fuel is cheap relative to its capital and it is not designed to follow load minute to minute. Gas and coal fill the middle. At the top are peaking units that may run only a few dozen hours a year, on the hottest and coldest days, and are paid accordingly.
Bar heights indicate relative marginal running cost and widths indicate relative share of capacity. The exact order differs by region, by fuel price and by hour.
The services that keep it stable
Energy is only part of what a grid buys. It also needs capacity held in reserve, equipment able to correct frequency within seconds, sources of reactive power to hold voltage steady, and a handful of plants capable of starting without any outside power at all. These are the ancillary services, and in most regions they are procured and priced separately from energy.
| Service | Time scale | What it does | Typical providers |
|---|---|---|---|
| Inertia | Instantaneous | Rotating mass slows the rate at which frequency falls after a loss of generation | Nuclear, gas, coal, hydro; synthetic inertia from batteries |
| Frequency regulation | Seconds to minutes | Continuous small adjustments up and down to hold 60 Hz | Batteries, hydro, flexible gas units |
| Operating reserves | 10 to 30 minutes | Capacity that is synchronized or can start quickly to replace a lost unit | Part-loaded thermal plants, combustion turbines, demand response |
| Voltage support | Continuous | Reactive power injected or absorbed to keep voltage within limits along the line | Generators, capacitor banks, synchronous condensers, inverters |
| Black start | Hours to days | Restarts the system from a complete shutdown without drawing outside power | Hydro units, designated diesel and gas turbines |
What happens when something fails
The system is planned so that it can survive the loss of any single large element without interrupting customers. Operators call this the N minus one criterion: with one component out, the remaining network must still carry the load within its limits. Meeting it means running lines below their maximum rating in normal conditions, which looks wasteful until the day it is not.
A generator trips offline
Within a second
Frequency begins to fall across the whole interconnection. The inertia of every spinning machine slows the decline.
Seconds
Governors on other plants sense the change and raise output automatically. Batteries respond almost instantly.
Ten minutes
Operating reserves are dispatched to restore frequency exactly and to rebuild the reserve margin for the next contingency.
If reserves are not enough
Operators shed load deliberately, rotating outages across neighborhoods. Controlled disconnection protects the system from an uncontrolled collapse that would take days to restore.
