megawatt.fun

A working roadmap for humanity

Build the power.
Climb the scale.

A civilization with more energy can do more. The work is to make that energy clean, dependable, affordable, and available to everyone.

Megawatt maps the path from today's energy system toward a Type I civilization, and the useful work we can start now.

Explore the plan Find a place to contribute
The climb in orders of magnitudePower, not a completion score
Humanity's path toward Type I A schematic logarithmic climb from about 20 terawatts today to 100 terawatts at K 0.8, 1,000 at K 0.9 and 10,000 at Type I. This illustration is not a forecast. We are here≈20.1 TW · K 0.730 100 TWK 0.8 1,000 TWK 0.9 10,000 TWType I

Every +0.1 on the scale requires 10× the power.

Our starting point

2024Latest available primary-energy row in the retrieved dataset

Average primary power

20.12 TW176,737 TWh ÷ 8,784 hours

Kardashev index

0.730A logarithmic energy-accounting proxy

Physical share of Type I

0.201%About 497× today's power at the target

2024 world primary energy, Our World in Data. Retrieved 21 September 2026. See the accounting caveat.

01 / The destination

Make the next rung tangible.

The scale gives us milestones. It does not prescribe an energy mix or promise a completion date.

K 0.8 / 100 TW

Make abundant power repeatable.

Scale clean supply, connect it, and make it dependable. Then repeat the whole deployment process across regions.

An editorial checkpoint, not an official Kardashev category or a forecast.

Power at this rung100 TW
Gap on the same accounting basis+79.88 TW
Multiple of the baseline4.97×
Physical power relative to the selected rung20.12%
20.12%

This gap is not a shopping list for electric generating capacity. Primary energy, useful electricity, and installed nameplate capacity are different quantities.

What would sustained growth imply?

A thought experiment with constant growth on a constant accounting basis. Efficiency, technology, and physical limits change the real trajectory.

0%5%
81 yearsaround 2105 from the 2024 baseline

Scenario only. No date here is a prediction or a committed deadline.

02 / The work

Build a system that can keep building.

These workstreams run together. Each has an owner, a dependency, and a result worth measuring.

ManufactureGenerateConnectBalanceDeliver useful work
Build clean generationTurn resources into dependable annual output.Deploy now

Build the locally appropriate mix of solar, wind, nuclear, geothermal and hydro. Improve existing plants and replace retiring high-emission supply. Keep research moving without making near-term progress depend on a breakthrough.

Who moves it
Developers, utilities, equipment makers, public and private capital.
Depends on
Sites, permits, equipment, a buyer and a viable connection.
Measure
Annual delivered TWh, lifecycle emissions, cost and reliability; report retirements and curtailment separately.
Connect supply to demandMake grid access part of the build plan.Priority to investigate

Expand transmission and distribution, improve interconnection studies, and use existing lines more effectively. Compare local connection lead times before choosing the binding bottleneck.

The IEA estimates that meeting national goals requires adding or refurbishing over 80 million kilometres of grids by 2040. This is a 2023 scenario estimate, not a Type I build requirement.

Who moves it
Grid operators, regulators, utilities and engineering contractors.
Depends on
Rights of way, transformers, skilled crews and coordinated studies.
Measure
Time to connection, commissioned transfer capacity, congestion and curtailment.
Make variable power dependableMatch energy to the hours people need it.Deploy + validate

Combine firm generation, transmission, demand flexibility, batteries and longer-duration storage. Design for difficult weather periods and equipment failures, not just the average day. Storage shifts energy and incurs losses; it does not create generation.

Who moves it
System operators, storage builders, controls engineers and flexible loads.
Depends on
Supply profiles, load forecasts, network constraints and trusted telemetry.
Measure
Unserved energy, duration of outages, forecast error, efficiency and delivered cost.
Inspect Megawatt's battery dispatch experiment
Scale the capacity to buildFactories, materials, finance and people.Expand now

Expand equipment manufacturing, train electricians and construction crews, qualify alternative suppliers, and recover materials through recycling. Standardize repeatable projects and shorten financing and permitting cycles without removing safety or community review.

Who moves it
Manufacturers, training providers, financiers and permitting authorities.
Depends on
Credible orders, material supply, skills and stable project requirements.
Measure
Equipment lead times, completed projects per year, build cost and safety performance.
Turn energy into better livesElectrify useful work and broaden access.Deliver now

Improve access to reliable electricity, electrify transport and heat where it works, and reduce energy wasted in buildings and industry. A more efficient system can provide more useful work while using less measured primary energy. That is progress even if K grows more slowly.

Who moves it
Communities, local utilities, builders, industrial operators and product teams.
Depends on
Affordable connections, reliable supply and equipment people can maintain.
Measure
Access, household energy burden, service reliability and useful output per unit of energy.
Earn the next order of magnitudeKeep the long horizon conditional.Research

At hundreds to thousands of terawatts, land, materials, ecosystem impacts and waste heat become central design questions. Study the physical limits and off-world options before treating Type I as an Earth-only construction target. Fusion and space-based power are research paths, not assumed contributions in this plan.

Who moves it
Energy researchers, Earth-system scientists, aerospace teams and public institutions.
Depends on
Demonstrated performance, complete energy balances and acceptable impacts.
Measure
Independently verified net energy, full-system cost, heat rejection and environmental consequences.

03 / The progress scorecard

More power. Uneven progress.

Track what is changing, not just what we want to build. These indicators have different dates and scopes; they do not add up to a single civilization score.

Sources reviewed

Next review due

Three global indicators and three US indicators. The US measures provide regional evidence about connections, outages and prices; they are not substitutes for global coverage.

EnergyWorld

Electricity generated

31,772 TWh

2025 · Annual generation

More output

+2.7% from 2024

More electricity was generated. The total alone does not establish cleaner energy, universal access, or dependable delivery.

Definition, comparison & source for Electricity generated

Measure. World electricity_generation series in the OWID energy dataset; annual electrical output, not primary energy or installed nameplate capacity.

Calculated annual comparison. World electricity_generation: 2023 = 29665.340; 2024 = 30930.230; 2025 = 31772.350 TWh, retrieved 2026-09-21.

Limit. Generation includes energy later lost or used within the energy system. It is not a measure of useful work delivered to homes and businesses, and must not be substituted into the primary-energy baseline above.

Published series, TWh
YearValue
202329,665.34
202430,930.23
202531,772.35

Next measurement. Pair new output with emissions, curtailment and reliability measures.

Our World in Data, Energy dataset
AccessWorld

People without electricity

655 million

2024 · People still unserved

Access rate stalled

Global access rate stagnated at 92%

The access gap remains large. The 2026 report describes a plateau in the global access rate, despite gains in some regions.

Definition, comparison & source for People without electricity

Measure. People lacking basic electricity access, as reported in Tracking SDG 7: The Energy Progress Report 2026, for 2024.

Source-reported direction. The 2026 release reports 655 million without electricity and describes the 2024 global access rate as stagnant at 92%.

Limit. A connection does not establish reliable service or affordability. No numeric year-over-year change is inferred by subtracting totals from different report editions, which may revise history.

Next measurement. Track affordable, reliable connections, especially in underserved regions.

World Bank and SDG 7 partner agencies, 2026 report
ClimateWorld

Energy-related CO₂

≈38.4 Gt CO₂

2025 · Annual emissions

Still rising

Around +0.4% in 2025, reported by IEA

Emissions grew more slowly, but still reached a new high. Slower growth is not an absolute reduction.

Definition, comparison & source for Energy-related CO₂

Measure. IEA's headline global energy-related CO₂ total, including fuel combustion, industrial processes and fugitive flaring.

Source-reported direction. The report gives nearly 38.4 Gt in 2025, about 145 Mt higher than 2024, and growth around 0.4%.

Limit. Not all greenhouse gases or all human-caused emissions. The growth rate comes from the same 2026 report; it is not calculated against the older 2025 report's unrevised total.

Next measurement. Measure absolute emissions alongside low-emission generation.

IEA, Global Energy Review 2026
Grid deliveryUnited States

Time to grid connection

>5 years

Built in 2025 · Median request-to-operation time

Getting longer

Increasing duration, reported by Berkeley Lab

Projects that reach operation face long waits. A shrinking queue alone is not proof that connection has become faster.

Definition, comparison & source for Time to grid connection

Measure. Median time from interconnection request to commercial operation for projects built in 2025, in regions with available data in Queued Up 2026.

Source-reported direction. The 2026 summary says the time to commercial operations is increasing and the median exceeded five years for projects built in 2025 where data are available.

Limit. A US completed-project cohort, not a global average or the expected wait for a new application. Withdrawn and still-active projects are not in this completion-time median; no exact year-over-year delta is claimed.

Next measurement. Separate study, equipment, permitting and construction delays.

Berkeley Lab and GridTracker, Queued Up 2026
ReliabilityUnited States

Time without power

≈11 hours

2024 · Annual interruptions per customer

Above historical norm

Nearly 2× the prior decade's annual average

Major events accounted for 80% of interruption hours in 2024. Resilience to extreme weather matters alongside routine service reliability.

Definition, comparison & source for Time without power

Measure. EIA's average non-momentary interruption duration per US electricity customer, including major events, from Electric Power Annual 2024.

Historical benchmark. EIA reports an average of 11 hours in 2024, nearly twice the preceding decade's annual average; major events account for 80%.

Limit. Compared with the 2014–2023 average, not just 2023. Major events dominate variation; this is not a weather-normalized trend or a worldwide reliability score.

Next measurement. Publish both all-event and normal-weather reliability measures.

EIA, Electric Power Annual 2024 / December 2025 analysis
Cost proxyUnited States

Residential electricity price

17.30 ¢/kWh

2025 preliminary · Nominal annual average

Higher nominal price

+5.0% versus 2024's annual figure

The nominal unit price increased. This is a price indicator, not a direct measure of household energy affordability.

Definition, comparison & source for Residential electricity price

Measure. EIA's US residential annual average retail electricity price: 17.30 cents/kWh in the preliminary 2025 series, versus 16.48 in the 2024 annual table.

Calculated annual comparison. Energy Explained cites 2025 residential 17.30 cents/kWh from February 2026 preliminary data; annual table 2.4 reports 16.48 for 2024.

Limit. 2025 is preliminary and may be revised. Not inflation-adjusted; bills also depend on usage, and affordability depends on income and other costs. This is not a global energy-burden measure.

Next measurement. Add household energy burden and inflation-adjusted prices before judging affordability.

EIA, Energy Explained and Electric Power Annual table 2.4Comparison source

Coverage still missing: a comparable global measure of outage duration, grid connection time, and household energy burden. Nominal electricity prices do not establish affordability.

Definitions & sources (JSON)Download indicators (CSV)

Manually reviewed snapshot. Review quarterly or when a source publishes a new edition; values do not update automatically. No forecasts are mixed into the observed indicators; preliminary estimates are labeled. A higher K is not a substitute for reliable, affordable, low-impact energy.

04 / Your part

Pick a constraint you can move.

You do not need to solve the whole energy system. Start with one real user, one measurable problem and one useful result.

Help operators make better decisions

Make dispatch claims testable.

Compare a simple battery schedule against forecast-based dispatch. Account for losses, reserve energy and the cost of being wrong.

Explore the dispatch experiment
  1. This weekRun the existing Go study, inspect a losing day and explain what the model leaves out.
  2. Evidence to earnA reproducible comparison with a baseline, feasible battery states and no look-ahead in the forecast policy.
  3. Who can use itBattery operators and engineers evaluating dispatch policies.

05 / Show the workings

A plan you can question.

Measured baseline, explicit calculations, and proposed priorities. No live global telemetry or claimed universal consensus.

How the baseline and milestones are calculated

The baseline uses the World 2024 primary-energy row from the OWID energy dataset: 176,737.094 TWh divided by the leap year's 8,784 hours gives 20.1203 TW. The retrieved dataset has no World primary-energy value for 2025. This is a dated snapshot, not a live estimate.

Using the modern continuous convention: K = (log₁₀(P in watts) − 6) / 10; equivalently, P = 10^(10K + 6). Type I is anchored at 10¹⁶ W. Decimal checkpoints are editorial, not additional named civilization types.

Primary-energy accounting includes conversion assumptions and is not the same as electrical output, useful work or dispatchable power. Do not mix accounting methods when comparing years or interpreting the gap. See the saved dataset metadata for the source method.

Growth scenarios use years = ceil(log(target / baseline) / log(1 + growth)), with the calendar result measured from 2024. Zero growth never reaches a higher target. Sustaining a constant rate for centuries is an assumption, not evidence.

Download the baseline and provenance (JSON)
Energy dataOur World in Data dataset and codebookThe scaleGray, The Extended Kardashev Scale (2020)The grid constraintIEA, Electricity Grids and Secure Energy Transitions (2023)Lessons from buildersMegawatt's energy-history study