Megawatts, gigawatts and the “homes powered” comparison

Convert MW and GW into an explicitly dated household-energy comparison while avoiding the common confusion between power, energy and capacity.

7 min read

“Enough to power a city” sounds clear until you ask which city, for how long and under what conditions. A megawatt is a rate of energy use or production. A household’s annual electricity bill records an amount of energy over a year. You cannot compare the two honestly without adding time and stating the assumptions.

This article works through that conversion. It is a scale illustration, not a promise that a data center takes power away from a fixed number of homes. It also is not a substitute for checking whether a campus figure means operating demand, IT equipment, connection capacity or a future plan.

Start with watts, then add time

EIA explains that watts measure power at a moment and watt-hours measure energy used over time. A kilowatt is 1,000 watts. A megawatt is 1,000 kilowatts. A gigawatt is 1,000 megawatts. These prefixes change the size of the unit, not the physical meaning of the measurement.

For constant power, energy equals power multiplied by time. One megawatt maintained for one hour produces or consumes one megawatt-hour. One megawatt-hour equals 1,000 kilowatt-hours. These are unit conversions, not estimates about a particular facility.

Think of a tap and a bucket. Power resembles the rate of flow; energy resembles the amount collected. A wide-open tap does not tell you how full a bucket becomes until you know how long it runs. Similarly, a planned electrical connection does not reveal annual energy consumption.

Choose a household figure and keep its date

EIA’s household-electricity FAQ reports that the average U.S. residential utility customer purchased 10,791 kilowatt-hours in 2022. That is an explicitly historical national average. It is not a current measurement for every household, and it should not be silently relabeled as a 2026 average.

The same FAQ warns that electricity purchases do not always equal total household consumption. Rooftop solar can reduce purchases from a utility. EIA’s household survey measures consumption differently. A comparison should say which figure it uses rather than mixing utility sales and household use.

Households also differ by climate, home size, heating equipment and other circumstances. A national average can make a large number easier to picture, but it cannot specify the needs of every local house. If a developer uses a local comparison, look for the local dataset and year.

There is another boundary to keep clear: household electricity is not all household energy. Fuels used for heating, transport or other purposes may sit outside an electricity comparison. “Homes powered” should not be expanded into a claim about all the energy used by the people living there.

Work through one megawatt

For this worked example, assume a constant one-megawatt electrical load for every hour of a non-leap year. There are 8,760 hours in that year. Multiplying one megawatt by 8,760 hours gives 8,760 megawatt-hours, or 8,760,000 kilowatt-hours.

Divide 8,760,000 kilowatt-hours by EIA’s 2022 average purchase figure of 10,791 kilowatt-hours per residential customer. The result is about 812 customer-years of electricity purchases. This is our arithmetic using EIA’s published baseline, not a site measurement or an EIA estimate of a data center.

The phrase customer-years is clumsy but useful. It shows that the comparison concerns an annual amount of electricity. Saying “812 homes” without the time period and baseline hides how the number was produced. A concise headline can still preserve those qualifications in the sentence below it.

The example assumes the load is constant. If equipment runs at different levels, use measured energy or a clearly supported operating profile instead. Do not treat an advertised maximum as an annual average simply because the multiplication is easy.

Scale the example to a gigawatt

One gigawatt is 1,000 megawatts. Under the same constant-load, non-leap-year assumptions, it represents 8.76 billion kilowatt-hours over a year. Dividing by the same 2022 baseline gives roughly 812,000 residential-customer-years of electricity purchases.

That larger number inherits every assumption from the smaller example. It does not become more certain because it sounds impressive. The household year remains 2022, the load remains constant, and the comparison remains about electricity purchases rather than every kind of energy use.

This is also not a statement that a one-gigawatt plant delivers that amount to homes. A generating plant’s actual output depends on operation. Grid losses and delivery constraints matter when discussing electricity delivered to customers. A capacity rating is not a guarantee of uninterrupted production at that rate.

For data centers, the equivalent caution is that planned campus capacity does not establish current consumption. A building can open in stages, or IT equipment can operate below its maximum. Read the announcement’s wording before applying any homes-equivalent calculation.

What the comparison cannot tell you

A homes-equivalent figure does not measure how household bills will change. Prices depend on costs, utility rules, supply conditions and other factors. Nor does it prove a data center will cause a particular outage. Reliability requires analysis of the actual network and resources.

It also cannot tell you the emissions associated with the electricity. The same annual quantity can come from different sources at different times. A household comparison is a scale aid. It is not an environmental assessment, a rate study or a grid-planning model.

Finally, comparing annual energy smooths out timing. Homes and data centers need electricity at particular moments, not only a sufficient total by December. A supply plan must address those moments. Annual matching and dependable delivery are related questions, but one does not automatically settle the other.

Keep the units attached when sharing the result

When passing along a comparison, carry its assumptions with it. “A constant one-megawatt load for a year” communicates something different from “a site with a one-megawatt connection.” The former states an operating assumption; the latter describes a limit or service arrangement whose actual use remains unknown.

Rounding should reflect the purpose of the calculation. The household average itself combines many different customers, so a long string of decimal places would imply more precision than the comparison deserves. An approximate result with a clear baseline is more useful than an exact-looking count of imaginary identical homes.

Be especially careful when the source number changes units partway through a story. A megawatt and a megawatt-hour do not become interchangeable because both are large. Check the suffix every time. If one number describes yearly energy and another describes peak demand, convert them only with a stated time or operating assumption. Otherwise keep them separate.

The same discipline works in reverse. If someone gives only a homes-equivalent claim, ask for the underlying electrical quantity before comparing it with another project. Without that original number and its definition, the comparison cannot be independently checked or updated to a different household dataset.

Why it matters

Large power figures deserve context, but context should make a claim clearer rather than louder. A transparent conversion lets readers reproduce the calculation, choose a different household baseline and see what changes. That is better than treating one dramatic number as a universal fact.

When you encounter a homes-powered claim, check the original power figure, its meaning, the time assumption, the household dataset and the year. If any are missing, the comparison is incomplete. The existing 100 MW explainer helps identify the original figure before this article’s conversion begins.

For quick reference, keep the chain in order: watts measure a rate; watt-hours measure an amount; an annual household baseline turns an annual amount into a comparison. Each step needs matching units. Once that chain is visible, megawatts and gigawatts become understandable without pretending every home or every campus behaves the same way.

Sources and further reading

Sources checked Oct 24, 2026. Worked examples and analogies are explanatory, not measurements of a real project.