International Energy Agency

What You're Getting

IEA provides global energy statistics: oil, gas, coal, electricity, renewables — supply, demand, trade, and CO2 emissions for 50+ countries (free tier) and all countries (licensed). Key datasets: World Energy Balances, CO2 Emissions from Fuel Combustion, Electricity Statistics, Oil Market Report (monthly), Gas Market Report. Annual frequency for most; monthly for oil market data. IEA data is authoritative for energy-to-emissions conversions.

Ingestion Strategy

The IEA does not offer a public API. Data access is through their website for free users or via bulk delivery for licensed subscribers. Free access covers OECD countries and basic aggregates in CSV/XLSX. For pipeline use, download the annual releases and track file updates manually or via web scraping.

import requests, io, pandas as pd

# Download IEA free data (requires browser session — best done manually)
# For licensed users, IEA provides SFTP or direct download links
# Community approach: cache downloaded files and process locally
def load_iea_csv(filepath):
    df = pd.read_csv(filepath, skiprows=1)  # IEA CSVs often have a header row to skip
    return df

Rate Limits & API Details

• No public API; data access is website-based for free users
• Licensed subscribers get bulk data delivery options
• Free tier: selected datasets for OECD countries at iea.org/data-and-statistics
• DataBrowser tool on the IEA website allows limited export (CSV/XLSX)

Schema Stability

IEA uses its own product and flow classification system (distinct from HS or SITC). Country names follow IEA conventions. Annual releases typically replace prior-year estimates with actuals. Energy units are standardized in TJ (terajoules) in the World Energy Balances — convert to desired units in your model layer.

Data Quality Gotchas

• No automation path for free data: without a license, pipeline automation requires web scraping (fragile) or manual download
• Product/flow taxonomy: IEA's classification (e.g. "Crude, NGL, feedstocks" vs "Petroleum products") differs from national energy agencies — document mappings
• CO2 calculation methodology: IEA uses specific emission factors; if comparing with IPCC national inventories, expect differences
• Preliminary vs final: monthly oil market data is preliminary; annual data is more reliable but lags 18–24 months

Tool Compatibility

• Python: manual download + pandas for free users; custom SFTP client for licensed bulk delivery
• R: manual download
• Not supported by standard ETL connectors
• For licensed bulk data: standard CSV/XLSX tools work

Benchmarking Approach

Global energy supply (TPES) should be ~580 EJ/year. OECD total final consumption should be ~220 EJ. CO2 emissions from fuel combustion: ~36 Gt globally. Cross-check electricity generation total for US against EIA figure — should be within 5%. Alert if annual release data (typically published in Q3) is not available by Q4.