Annexe · sources, methods and downloads
How the numbers on this site are made
Where every figure comes from, how it is calculated, what it can and cannot tell you, and the files behind it. Energy data runs to
A. Principles
- Levers first, outcomes second. Policy, technology and business indicators lead; demand, prices and supply chains show the impact.
- Strong signals move a status; weak signals only show direction. A challenger technology moves up the scorecard, or a target changes status, only on evidence: costs reached, capacity installed, a rule enforced, money spent. Announcements, drafts and memoranda are logged as weak signals.
- India and the world side by side, on the same axes and units.
- Trend before level. Ten-year changes and compound growth come before single readings, and every ten-year window ends at the latest year in the data.
- Authentic sources only. Every record links to an official, intergovernmental or primary source, or to a recognised data provider labelled as such. News reports are used to find the underlying release, never cited. Opinion and sources known to be unreliable are excluded.
B. Sources
| Dataset | Covers | Latest | Refreshed |
|---|---|---|---|
| Our World in Data energy dataset, compiled from the Energy Institute Statistical Review of World Energy 2026 and Ember | Primary energy, electricity and production by fuel for India, the world, China, the US and the EU | Weekly check; the Statistical Review is annual (next due mid-2027) | |
| US Energy Information Administration spot prices, via Frictionless Data | Brent crude and Henry Hub gas, monthly | Weekly | |
| World Bank Pink Sheet | European gas, LNG landed in Japan, Australian coal, monthly | Weekly; the World Bank updates early each month | |
| US Bureau of Labor Statistics consumer price index | Used to restate dollar prices after inflation | Weekly | |
| Curated records | Targets, policies, investment, deals, costs, technologies, levers and risks | Each record carries its own date and link | Reviewed by the editor; weekly to quarterly |
Source tiers for business and technology records. Tier 1 is official or intergovernmental: governments and regulators, the IEA and IRENA, auction agencies (SECI, NTPC), stock exchanges and the SEC, and companies' own releases about their own deals. Tier 2 is a recognised data provider or research body (BloombergNEF's free summaries, the Fusion Industry Association, CEEW, World Nuclear Association, PRS Legislative Research), always labelled on the page and never blended with Tier 1 figures. Policy targets and retail prices accept official government sources only.
C. Methods and units
Primary energy. Measured in terawatt-hours (1 EJ = 277.8 TWh). From its 2026 edition the Energy Institute uses the physical energy content method: coal, oil and gas count by the heat they release when burned; solar, wind and hydro count as the electricity they generate; nuclear, geothermal and bioenergy keep a thermal conversion. Earlier editions used the substitution method, which counted non-fossil electricity as the fossil fuel it would replace, inflating it about 2.5 times. The switch lowers the measured size of solar, wind and hydro, raises fossil fuels' share by several points (the world's fossil share reads about 86% in 2025, against about 80% under the old method), and is applied to the whole history, so trends stay comparable. Our World in Data's explainer.
Electricity. Generation and demand in TWh; shares are of electricity generated in the year, never of installed capacity.
Useful energy. The Prices page divides each price by the energy in one unit and by a typical device efficiency (petrol engine 25%, diesel engine 30%, electric car from the socket 80%, gas stove 60%, induction hob 85%, combined-cycle gas plant 55%). It covers the energy cost only, not the cost of buying the car, stove or plant.
Prices and inflation. Dollar prices are monthly averages of the price paid at the time ("nominal"). Wherever a page says "after inflation" or "in
Energy content of benchmarks. One barrel of crude is taken as 1.7 MWh, one MMBtu as 0.293 MWh, and one tonne of Australian coal (6,000 kcal/kg) as 6.98 MWh.
Import shares. Net imports divided by consumption, in energy terms: (consumption − production) ÷ consumption. India's oil ministry measures by volume and financial year, so its figures differ slightly (for 2023-24, about 85% for crude oil and 46% for gas).
Growth rates. Compound annual growth over ten years ending at the latest year. A high rate from a small base can add less energy than a low rate from a large one, so pages show the absolute change beside it.
Target status. Met when reached. Otherwise the pace achieved in the latest year is compared with the pace still needed: On track at 90% or more, Behind at 60–90%, Off track below 60%. A status set by hand always shows its reason.
Direction of demand. The latest year's change against the ten-year average: Accelerating above 1.3 times, Slowing below 0.7 times, Falling if negative. Where a source reports a newer year than the dataset, its direction is shown and linked.
D. Reading the figures safely
Small i buttons beside figures explain how each could be misread. The most common traps:
- Capacity is not generation. India's non-fossil sources passed half of installed capacity in 2025 but made about a quarter of its electricity.
- Shares of electricity are not shares of energy. Solar made over 9% of India's electricity in 2025 but under 2% of its primary energy.
- Primary energy overstates fuels that waste heat. A TWh of coal burned for power yields about a third of a TWh of electricity.
- Costs of new supply are not system costs. Auction tariffs and levelised costs compare new plants; they leave out integration costs and the low running cost of plants already paid for.
- Estimates get revised. Current-year investment figures and early capacity counts change; agencies also differ (IEA and IRENA count 2025 renewable additions at about 800 GW and 693 GW).
- One source per comparison. Where two sources disagree, the page names both and does not mix them in one chart.
E. Data downloads
Every dataset behind the site, free to reuse with attribution to the original source.
| File | What it holds | Source and licence |
|---|---|---|
| owid_energy.csv | Annual energy and electricity by fuel for India, the world, China, the US and the EU | Our World in Data (Energy Institute, Ember), CC BY 4.0 |
| prices.csv | Monthly benchmark prices, at the time and after inflation | US EIA and BLS (public domain); World Bank (CC BY 4.0) |
| scorecard.json | Disruption scorecard stages | Editorial |
| hypotheses.json | Key hypotheses with direction of travel, signal, key figures, indicators and evidence chart settings | Editorial |
| targets.json | National energy and climate targets with progress, pace and status | Official sources, linked per record |
| commitments.json | Headline commitments by economy | Official sources, linked per record |
| investment.json | Energy capital spending by economy and sector | IEA (Tier 1) and BloombergNEF (Tier 2), linked per record |
| venture.json | Venture capital for emerging themes, by theme and region | IEA (Tier 1) and Fusion Industry Association (Tier 2) |
| market_signals.json | Auction tariffs, orders and other economic signals | Tier 1 or Tier 2, linked per record |
| deals.json | Large deals, auction awards and listings by stage | Filings, auction results and company releases |
| cost_race.json | Challenger against incumbent costs, with the verdict worked out from the figures | IRENA, IEA, SECI, filings (Tier 1) and CEEW (Tier 2), linked per record |
| cost_trends.json | Cost trends and headline shares for clean technologies | IRENA, IEA (Tier 1) and BloombergNEF (Tier 2) |
| disruptors.json | Readiness of unproven technologies, with dated evidence | Regulators, official releases and company results, linked per record |
| watchlist.json | Milestones due in the next 12 to 24 months and who set each date | Linked per record |
| demand_levers.json | Levers pushing each fuel's demand up and down, with the IEA's reading of the latest year | Editorial; IEA |
| retail_prices.json | India's retail petrol, diesel and LPG prices (latest and year-wise averages since 2015-16) and household tariffs | Oil ministry, Parliament answers (PPAC data), Central Electricity Authority |
| end_uses.json | Energy content and device efficiency used to work out useful-energy costs | Editorial assumptions, shown on the page |
| price_risks.json | Exposure of each price to five kinds of risk | Editorial |
| build_profiles.json | Size and build time of each kind of energy project | Editorial, typical values |
| supply_chains.json | Supply-chain readiness link by link, for India and the world | Editorial |
| chain_risks.json | Risks that cut across supply chains | Tier 1 or Tier 2, linked per record |
| policies.json | Policy instrument tracker | Official sources, linked per record |
| weak_signals.json | Weak signals log | Official sources, linked per record |