Turkey Geothermal Performance Report 2026 — Summary

Turkey operates the world's fourth-largest geothermal power fleet: 1798.3 MWe across ~70 plants, generating ~8.7 TWh per year — virtually all of it in two grabens of the Menderes Massif. This report unifies the fleet's public performance data for the first time: plant-by-plant capacity factors, the feed-in-tariff (YEKDEM) exit calendar, and the collapsing spot price they will face.

Five findings

1. Output is flat, efficiency is eroding. Fleet capacity factor slid from 61.2% (2021) to 57.1% (2025) while installed capacity kept growing.
2. A two-speed fleet. The 45 plants still under the YEKDEM feed-in tariff have held a collective ~63% capacity factor for five years. Every cohort that has left the support scheme runs lower — and declines through its exit window.
3. The cleanest natural experiment: the 6 plants whose support ended in 2026 were stable at ~47% for four straight years, then dropped 3.7 points in their first unsupported year — the same year spot prices collapsed. Exiting cohorts were already weaker before exit (selection effect), so we read this as economics amplifying, not creating, the gap.
4. The YEKDEM cliff: 213.4 MWe leaves the USD-denominated tariff at the end of 2026; a cumulative 1057.8 MWe (~59% of the fleet) by end-2031. At 2026 price levels this implies a revenue gap on the order of $60-105M per year for each exiting vintage (scenario, not a forecast — see methodology).
5. The price scissors: the YEK-1 tariff guarantees 105 USD/MWh; the 2026 day-ahead price has averaged 39 USD/MWh (May 2026: 13 USD/MWh), driven by spring solar/hydro abundance in exactly the hours baseload geothermal cannot avoid.

Data & method, in one paragraph

Production is settlement-grade monthly data (EPİAŞ transparency platform) for Jan 2020 – Jun 2026; our fleet total matches the official national geothermal total in 78 of 78 months (0.00% deviation). Capacity factors use date-ranged installed capacity with unit-acceptance dates curated from archived press and regulator filings. Anomalous months (CF>105%, a real winter effect in binary plants) and the first six months after commissioning are excluded. The seven most extreme plant findings were independently recomputed from raw hourly archives before publication. Everything — dataset (CC BY 4.0), per-plant provenance and full methodology — is open at the links below.

Sources: EPİAŞ Transparency Platform (production, prices; used with attribution), EPDK YEK lists 2016-2026, JESDER, Global Energy Monitor (coordinates/technology, CC BY 4.0). Not investment advice. Found an error? Tell us — we recompute from source and publish corrections.