Who decides what, from a light switch to a nation?

Zoom from one home to a whole nation, and sort who decides what at each scale

F1 of Energy Systems, Step by Step: zoom from one home to a fictional nation, sort who decides, what they decide and what physics a model needs at each of five scales, and see what passes between scales.
  • Foundations · L1
  • 15 min guided · about 10 min practice
  • After lesson 0

The idea in one table: five scales, five sets of decisions

Scale Who decides What they decide What a model needs
S0 Device and plant The owner of a home or plant What to install at one site, and how to run it hour by hour Device ratings, efficiencies, one meter
S1 Microgrid and community A campus owner or a rural co-op Size, and which loads stay on when it islands A small network that can run alone
S2 Regional bulk grid A grid operator, overseen by regulators Which plants run in its region, reserves, lines, connections Power flow on lines; frequency and inertia for fast events
S3 City and region A city or county Building codes, heat, transit, land, feeder upgrades Feeders, heat networks, links between sectors
S4 Nation A national government and regulators Targets, instruments, infrastructure, trade Regional energy balances, main corridors
  • Electricity a U.S. home buys from its utility, as average power: about 1.2 kW (2022).
  • The Southwest Power Pool's highest hour of 2025: 54.4 GW.
  • The Lower 48 states' record hour: 759 GW (29 July 2025).
  • Units: 1 kW = 1,000 W (watts); 1 MW = 1,000 kW; 1 GW = 1,000 MW.

The scales overlap. Signals and limits usually pass down, such as a utility's tariff to a home; profiles and needs pass up, such as a home's load to the grid operator.

With JavaScript on, you work through this step by step: a zoom through five scenes, five sorts and one question about a campus.

Home: EIA's 10,791 kWh a year bought per residential customer (2022 data) ÷ 8,760 hours.
Southwest Power Pool: its 2025 peak, from EIA-930.
Lower 48: EIA's preliminary record.
Fictional: Prairie Town, Prairie Campus, the Plains Grid, Prairie City and Prairieland; their sizes are illustrative. Prairie Town's 450 MW is the highest hour of its illustrative year.

What you just did, as diagrams

A scale
A view of the system with its own decision makers, decisions and physics, not just a size on the map. A microgrid (S1) can even island.
Who decides
An owner (S0), a campus or co-op (S1), a grid operator (S2), a city (S3), a nation (S4).
Between scales
Prices, tariffs, limits and targets usually pass down, such as a utility's tariff to a home; load profiles, flexibility and needs pass up.
Local optimum
The best plan for one scale, a local optimum, holds only inside its boundary; a limit from the scale above can change it.
Go deeper · AnalystReal organisations straddle scales

The five scales are a lens for questions, not an organisation chart.

  • A co-op can sit on two scales. It can own a campus microgrid (S1) and buy power on the regional grid (S2).
  • One site can matter to a whole grid. A large factory or data-center campus is one site (S0), yet large enough to change regional planning (S2): Analyst lesson A6.
  • One question can span two scales. Where new solar goes is a city question (land, S3) and a grid question (connection, S2).
  • Regulators reach across. One regulator can oversee a utility’s grid (S2) and the tariffs its city customers pay (S3): Analyst lesson A4.
Go deeper · ModelerLinking models across scales: a preview

A model of one scale takes the arrows from other scales as given data, not as its own decisions: the Given bin of lesson 10. The campus question, in words:

  • Decision: how much the campus exports in each hour.
  • Goal: the cheapest campus plan.
  • Rule passed down from the grid: in every hour, export (MW) ≤ the feeder’s export limit (MW).

When two scales’ models pass numbers back and forth, a modeler must check whether their answers settle; they may not. Modeler lesson M5 adds the feeder’s limit to a campus design; capstone C1 asks who pays.

Go deeperWhat this lesson leaves out
  • Grid physics: how a grid stays balanced is lesson 3; wires and outages are F6.
  • Rules and money: who pays at each scale is Analyst lesson A4, and A6.
  • Countries: institutions differ by country, and the course's regional cards and F10 compare them.
  • Whole-system effects: how one change ripples across every scale is F2.

Check yourself

Practice, about 10 minutes: three myths, then six situations to place on the ladder.

Myth or fact? “A scale is just a size on the map.”

  • Myth
  • Fact
Answer

Myth. A scale is set by who decides what and which physics a model needs. Prairie City lies inside the Plains Grid's area, yet the city and the grid operator decide different things.

Myth or fact? “The cheapest plan for one campus is also the best plan for the grid it sits on.”

  • Myth
  • Fact
Answer

Myth. A local optimum holds only inside its boundary; the grid weighs everyone's costs and limits, and the two answers can differ.

Myth or fact? “One household's choices are too small to matter at the national scale.”

  • Myth
  • Fact
Answer

Myth. One home is small, but millions of homes' profiles add up to the demand that grids and nations plan for: needs pass up the ladder.

A rural co-op decides which loads stay on when its microgrid runs on its own in an outage. Which scale?

  • S0 Device and plant
  • S1 Microgrid and community
  • S2 Regional grid
  • S3 City and region
Answer

S1: load priorities when a microgrid islands are the microgrid's own decision.

An operator calls on reserves after a large plant trips. Which scale?

  • S0 Device and plant
  • S1 Microgrid and community
  • S2 Regional grid
  • S4 Nation
Answer

S2: the regional operator holds reserves to keep supply and demand balanced in its area (lesson 3).

A county requires new homes to be wired for heat pumps and EV chargers. Which scale?

  • S0 Device and plant
  • S2 Regional grid
  • S4 Nation
  • S3 City and region
Answer

S3: here the county sets the building code, and that shapes demand for years.

A ministry decides how many days of oil imports to keep in emergency storage. Which scale?

  • S4 Nation
  • S3 City and region
  • S2 Regional grid
  • S1 Microgrid and community
Answer

S4: emergency fuel stocks are a national security decision.

Which physics might a regional grid model need that a one-home scheduling model usually leaves out?

  • A battery's state of charge
  • Power flow on lines and the grid's frequency
  • The efficiency of a heat pump
Answer

Power flow on lines, and the grid's frequency when a study looks at fast disturbances. A home scheduling model needs its devices, not the network.

Which of these is usually sent down, from a larger scale to a smaller one?

  • A home's hourly load, sent to the grid operator
  • A city's demand forecast, sent to regional planners
  • A utility's tariff, sent to its customers
Answer

The tariff: prices, limits and targets usually pass down; loads, forecasts and needs pass up.

Numbers, sources and what’s next

Numbers to remember

Number What it means Source
1.2 kW Electricity bought per U.S. residential customer, as average power: 10,791 kWh a year ÷ 8,760 hours (S0). Power made on site, such as rooftop solar, is not counted. EIA, FAQ: How much electricity does an American home use? · 2022 data, read 3 November 2026
54.4 GW The Southwest Power Pool's highest hour of demand in 2025 (S2). EIA, Hourly Electric Grid Monitor (Form EIA-930) · 2025 data, read November 2026
759,180 MW The Lower 48 states' record hour of demand, 29 July 2025 (S4). EIA, Today in Energy · 5 August 2025, preliminary data

Next

F2 · How does everything connect? Stocks, flows, feedback and the whole-system map

In the real world

  1. Grids: the Lower 48 states have three main interconnections, which rarely trade power with each other.
  2. Operators: within them, balancing authorities, mostly utilities, run the day-to-day balance of their own regions.
  3. Cities, states, nation: cities and states shape demand and siting; national rules set the frame.

Sources and further reading

Spotted an error, or stuck? Write to hlli@ksu.edu. Corrections are made in place and dated here.

Text licensed CC BY 4.0, code MIT: teachers and students may reuse and adapt it, with credit to Honglin Li.