Grid-building revolutions succeed on balance sheets and institutions, not engineering alone — and 121 of 180 countries in the 2022 table still sit below 4,000 kWh per person

Asked (summary)

How did every country in the world go through — or not yet go through — its grid-building revolution? Test, rather than assume, the hypothesis that generation is primarily an engineering problem, transmission a coordination problem, distribution an operations problem, and universal reliable electricity ultimately a capital-formation and institutional problem — with a global stage map, historical revolution periods, bottleneck cohorts below 4,000 kWh/person/year, archetypes, 20 named case studies, threshold analysis at 500/1,000/2,000/4,000 kWh, failures, international-institution mapping, and a grand synthesis.

This first-edition atlas maps 180 named countries and territories from the 2022 GlobalPetrolPrices consumption table (169 sovereign states plus territories; an older 212-entry Worldometer table and a 215-country World Bank access page provide cross-checks), 943 extracted historical country-episodes spanning roughly 1880–2025 across 111 country labels, a 30-row normalized case-study table covering all 20 required countries, and a peer-reviewed 178-country infrastructure inventory (1980–2017). Stages are a transparent provisional rubric built on consumption intensity plus episode evidence — not on access rates alone — and unknowns are marked as unknown.

Caveats up front: consumption cohorts are analytical markers, not welfare requirements or proof of causality; high access does not imply high reliability; many extracted date ranges combine different episodes and are labelled approximate; the infrastructure inventory warns its distribution-line lengths are often regression estimates; and the 2024 access ranking claims 215 countries but displayed only its first 50 rows as fetched.

Global stage map — provisional consumption-intensity cohorts, 2022

Pre-grid <500 kWh/person Early build 500–999 Building 1,000–1,999 Consolidating 2,000–3,999 High-energy ≥4,000 Unknown — no 2022 row

Hover or tap a country for its 2022 per-capita and total consumption and how many historical episodes the evidence bank holds for it. Colour is 2022 consumption intensity only; it deliberately does not claim reliability. Grey countries have no row in the 2022 named-country table (the older 212-entry table covers more territories). Territories are flagged in the table below.

Stage counts (2022 table): 42 countries below 500 kWh/person, 20 at 500–999, 25 at 1,000–1,999, 34 at 2,000–3,999, and 59 at or above 4,000 — so two-thirds of listed countries are still under the 4,000 marker. globalpetrolprices.com
The spread is 3,600-fold: Iceland leads at 50,602 kWh/person and Sierra Leone trails at 14; China consumes the most in total (8,349 billion kWh) and Kiribati the least (0.03). globalpetrolprices.com
Tables disagree: the older Worldometer/EIA table puts China at 3,963 kWh/person versus 5,912 in the 2022 table — vintage matters, so comparisons here prefer the explicit 2022 table and note the older one. srv1.worldometers.info
Access is not reliability: the World Bank access series counts a connection from surveys, censuses or provider reports, models gaps from cross-country trends, and assumes developed countries universal — none of which measures hours of stable supply. ourworldindata.org
The wires themselves: the 178-country inventory estimates 4.7 million transmission circuit-km and about 96 million distribution circuit-km worldwide in 2017 — a 20:1 ratio that makes distribution, the operations layer, the overwhelming bulk of the physical grid — while warning distribution lengths are often regression estimates. ncbi.nlm.nih.gov

When revolutions happened — 943 evidence episodes by start decade

Each bar counts extracted country-episodes (a dated decisive event in a source) whose start year falls in that decade; 725 of 943 episodes carry a usable four-digit start year. The bulge from the 1990s onward reflects both real reform waves (unbundling, IPPs, access programs) and the survivorship of recent documents. Episodes are candidates, not definitive national revolutions.

Does the hypothesis hold? A qualified yes

What the 943 episodes repeatedly show

  • Transmission is coordination-heavy. Britain needed the 1926 Act and a statutory Central Electricity Board to force a national 132 kV grid on fragmented private suppliers; the TVA was praised in India precisely for solving a seven-state collective-action problem no single state could; the France–Spain HVDC link took an intergovernmental agreement, an EU grant and an EIB loan covering ~82% of cost.
  • Distribution is operations-heavy. Buenos Aires' Acuerdo Marco tied subsidies to actual service and drove losses in poor neighbourhoods below 10% with collection above 97%; Delhi privatized distribution on loss-reduction bids; Bangladesh's REB scaled to 500,000 new customers a year on procedure and metering discipline.
  • Universal reliable supply is capital-and-institutions. The UK's nationalized industry absorbed 8–11% of all national fixed investment for two decades; India devoted nearly one-fifth of all planned expenditure to power for thirty years; sovereign guarantees, World Bank covenants and tariff institutions recur in nearly every successful case.

Where the neat labels break

  • Generation is not just engineering. Ghana's Akosombo dam pencilled only because a thirty-year VALCO smelter contract at 2.625 US mills/kWh de-risked demand; Morocco's Jorf Lasfar IPP needed World Bank and export-credit political-risk guarantees plus take-or-pay clauses.
  • Distribution needs capital and coordination too: Orissa's privatization foundered when the state diverted the earmarked subsidy; Egypt's municipal distribution left $76 million unpaid public accounts by 1975.
  • External shocks can overwhelm sound institutions: conflict, fuel-import exposure and macro crises recur as failure triggers in the episode bank independent of utility design.
  • Verdict: the hypothesis survives as a statement of the binding constraint at each layer, not an exclusive one. Every durable success combined system-wide planning, long-duration or state-backed capital, explicit risk absorption, anchor loads, and predictable revenue recovery.

Twenty required case studies, thirty evidence rows

Dates are approximate/provisional: extraction aggregates multiple episodes per country, so some min–max spans (e.g. Brazil 1897–2047, which mixes an early episode with a concession end-date) are extraction artefacts, framed here by the event text instead. Subnational and cross-border variants are kept because they carry distinct mechanisms.

Thresholds, archetypes and the sub-4,000 bottleneck cohort

International institutions and the proposed global grid-asset institution

All countries in the 2022 consumption table

CountryStage cohortkWh/person 2022Total bn kWhEpisodes in bankSource

Data-gap register and what 10–20 years can buy

Method: stage cohorts computed from the GlobalPetrolPrices 2022 named-country consumption table (180 parsed rows incl. 6 territories, flagged; the page states 169 countries); cross-checked against the older 212-entry Worldometer/EIA table and a 2024 access page claiming 215 countries but displaying 50 fetched rows. Historical evidence: 943 extracted country-episodes (111 country labels, 725 with usable start years, mined for mechanisms, ~1880–2025) and a 30-row normalized case-study table covering the 20 required countries; dates approximate where episodes were aggregated. Infrastructure stocks from a peer-reviewed 178-country inventory, 1980–2017; its distribution lengths are often regression estimates. kWh/person measures annual electricity consumption per capita. Country names normalized to Natural Earth spellings; countries without a 2022 row are drawn grey as unknown. Cut for space: per-country episode texts beyond the case cards, and the older table's full 212 rows.

This report was generated automatically by Keenable SELECT at a user's request, from publicly available web sources linked herein. Keenable does not review, verify, or endorse its contents and makes no representation as to accuracy, completeness, or timeliness; AI-based extraction may contain errors. Nothing in this report is investment, legal, financial, or other professional advice. All trademarks and referenced content remain the property of their respective owners; no affiliation or endorsement is implied. To report an error, rights concern, or request removal: legal@keenable.ai.

Keenable SELECTAsk your own question
Made with Keenable SELECT