Supporting materials

Downloads

Methodology

Electricity generation, imports, demand and emissions

Annual data from 2000 to 2024 covers gross generation, sourced from the Energy Institute’s Statistical Review of World Energy, the Energy Information Administration (EIA), Eurostat, IRENA, and national sources such as China’s National Energy Administration the UK’s Department of Energy Security and Net Zero, and Chile’s Coordinador Eléctrico Nacional. 2025 data represents estimates of gross generation based on monthly generation data. This estimate is derived by applying absolute changes in monthly generation to the most recent annual baseline. For some countries, such as Ethiopia or Uzbekistan, dedicated annual data sources are used to estimate generation in 2025. A full list of sources for monthly and yearly data and more information about the dataset can be found here.

Net imports from 1990 to 2024 are taken from the EIA and Eurostat, with recent 2025 data estimated in the same manner as generation. Demand is calculated as the sum of generation and net imports, and validated where possible against published direct demand figures. Because it uses gross generation and does not include transmission and distribution losses, it will tend to be higher than end-user demand.

Monthly data is collected for 87 countries from over 70 sources, including national transmission system operators and statistical agencies, as well as data aggregators such as ENTSO-E. In some cases, data is published on a monthly lag; in such cases, recent months are estimated based on Ember’s own generation forecasting model.

Both annual and monthly data is often reported provisionally and subject to revision. Every effort has been made to ensure accuracy, and where possible we compare multiple sources to confirm their agreement.

Bioenergy has typically been assumed (by the IPCC, the IEA, and many others) to be a renewable energy source, in that forest and energy crops can be regrown and replenished, unlike fossil fuels. It is included in many governmental climate targets, including EU renewable energy legislation, and so Ember includes it in “renewable” to allow easy comparison with legislated targets. However, we recognise the IPCC reported lifecycle carbon intensity of bioenergy is significantly higher than other renewables and nuclear, and this is incorporated into our power sector emissions estimate. More information about Ember’s classification of electricity sources can be found in the full methodology for Ember’s Yearly Electricity Data under “Fuel Types”.

References to CO2 emissions in this report use CO2 equivalent emissions, which include emissions from other greenhouse gases such as methane (CH4). Power sector emissions are also based on the methodology from Ember’s Yearly Electricity Data.

Solar generation and distributed solar estimates

Official generation data for solar power often underreports generation from distributed solar, or “behind-the-meter” systems. Where possible, distributed solar generation is included in Ember’s electricity data and the data used in this report. For countries where it is not directly reported, generation from distributed systems is estimated using various approaches. For example, India’s official solar generation data reported by the Central Electricity Authority does not include output for all small-scale systems. For India, solar generation is scaled based on the ratio between capacity with generation reporting and capacity without generation reporting (mostly small-scale systems). A performance penalty is applied for performance in small-scale distributed systems, as they tend to have a lower capacity factor than utility-scale systems.

Further information on scaling approaches, behind-the-meter solar coverage, and detailed methodology breakdowns by country can be found in Ember’s monthly and yearly electricity dataset.

Solar and wind capacity

Solar and wind capacity deployment data is sourced from Ember’s Monthly Wind and Solar Capacity Data, which tracks monthly deployment for 25 countries, covering 93% of solar and 92% of wind capacity installed globally as of 2025. Global capacity additions are estimated using monthly capacity data from national sources through end-2025. Global deployment data for 2025 requires estimates for data not covered in country-level capacity tracking. For solar, estimates for remaining countries are derived from analysis of Chinese solar PV module export data. For wind, the total deployment reported by available countries is scaled up to global values using the ratio of their combined capacity to total global installed capacity, based on previous years with full coverage.

Solar capacity is reported in either GW(AC) or GW(DC) depending on reporting conventions and data sources. In the report, it is clearly marked as either AC or DC. GW(DC) values are the nameplate capacity, reflecting the maximum potential panel output. Global additions and solar export figures in the report are typically reported in GW(DC). For some country-level figures, such as India’s 2025 additions, the report uses GW(AC) data, which represents the maximum output capacity of a power plant at the grid connection point. DC capacity is typically around 1.2 to 1.3 times larger than AC capacity.

Further information on the dataset, its sources, and a detailed methodology can be found on the dedicated page on Ember’s website.

Calculation of temperature impacts on electricity demand

Temperature impacts on electricity demand were estimated using regression analysis applied to population-weighted cooling degree days (CDD, above 22°C) and heating degree days (HDD, below 18°C), derived from ERA5 hourly temperature data accessed via the EU’s Climate Data Store. The analysis was conducted for 38 countries and regions covering 86% of global electricity demand, with results scaled to represent global totals. Monthly electricity demand, sourced from Ember’s Monthly Electricity dataset, was normalised against a 12-month trailing average to isolate temperature-driven variations from structural trends such as economic growth or electrification. Temperature anomalies were calculated relative to a 2015-2024 baseline and used to derive absolute demand impacts in TWh, enabling a clearer assessment of underlying structural changes in electricity demand.

Battery storage

The share of new solar generation that can be shifted with new battery capacity is estimated based on Ember’s yearly electricity data for solar generation and battery capacity installations measured in GWh from various sources. The year-on-year increase of solar generation in 2025 compared with 2024 in GWh is divided by the number of days in 2025 to produce the increase on the average day. The battery capacity installed in 2025 is then divided by the daily growth in solar generation, assuming one full cycle per day, as an order of magnitude estimate of how much of the new daily solar generation can be absorbed by the battery capacity on the average day. No additional assumptions are made on battery operation such as depth of discharge.

Representative global solar profile

The estimated hourly share of solar power in global electricity demand uses representative hourly profiles from large electricity markets with available hourly data like the US, EU, India, and Brazil. The derived profile was scaled to match the monthly average at global level.

EV demand and oil displacement

Electricity demand growth from EVs is estimated by multiplying changes in EV stock, disaggregated by vehicle type, by reference electricity consumption values per vehicle type. Vehicle types covered include passenger cars, buses, trucks, and vans, across both battery electric vehicles (BEVs) and plug-in hybrid electric vehicles (PHEVs). Historical reference data is sourced from the IEA’s Global EV Data Explorer. 2025 demand growth is estimated using 2025 EV sales data published by BloombergNEF. Oil displacement estimates are based on historical IEA values, assuming the ratio between EV stock and displaced oil remains constant by vehicle type between 2024 and 2025.

LNG volumes and electricity generation from gas equivalent

LNG volumes (metric tonnes) are converted to electricity generation using the IEA standard net calorific value of 48.6 GJ/tonne (LHV basis). A 50% plant efficiency factor is applied, representing a modern combined cycle gas turbine (CCGT), yielding a conversion factor of 6.75 MWh per tonne of LNG.

Other data sources

This report makes use of a variety of datasets curated by Ember, including data on exports of Chinese solar PV modules. A full methodology for this dataset can be found here.

All of Ember’s datasets, including a variety of data tools for exploring the data, are available on Ember’s data page.

Acknowledgement

Lead Author

Nicolas Fulghum

Other authors

Katye Altieri, Kostantsa Rangelova, Wilmar Suarez

Data visualisation

Chelsea Bruce-Lockhart, Lauren Orso, Jivan Zhen Thiru

Project Manager

Hannah Granados Smith

Editor

Raul Miranda

Communications

Led by Rini Sucahyo with support from Alison Candlin, Ardhi Arsala Rahmani, Burcu Unal Kurban, Claire Kaelin, Eli Terry, Eva Mbengue, Hannah Broadbent, Izabela Urbanska, Rashmi Mishra, Reynaldo Dizon, Rocío Rodríguez Almaraz, Sachin Sreejith, Shiyao Zhang, Taiki Asato, Tito Das.

Other contributors

Alnie Demoral, Beatrice Petrovich, Biqing Yang, Chris Rosslowe, Dave Jones, Dinita Setyawati, Duttatreya Das, Euan Graham, Giang Vu, James Blackwell, Josie Murdoch, Leonard Heberer, Libby Copsey, Matt Ewen, Muyi Yang, Neha Rajput, Richard Black.

Peer reviewers

Bryony Worthington (Ember), Hannah Ritchie (Our World in Data), Harry Benham (Ember), Kingsmill Bond (Ember), Nathaniel Bullard (Business Climate Pte. Ltd), Xunpeng ‘Roc’ Shi (UTS/ISETS).

Cover image

The cover of this report was designed by Reynaldo Dizon.