Utilizing Geothermal Waste Heat for District Cooling Using Absorption Chillers: A Pilot Project in Szeged, Hungary (EGC 2025) + Geothermal Rising data-center RFI

Provenance: Conference paper from the University of Szeged geology group, InnoGeo Research and SZETÁV Ltd. (the city’s district-heating utility). Design phase — no operating data yet. The paper is internally ambiguous on the exact drive temperature (abstract cites 95 °C production water; conclusions cite 40–60 °C return water). Retrieved via research agent; PDF text extracted locally.

The Szeged system

27 wells (9 production, 18 injection); ~95 °C production water from 2,000 m Upper Pannonian sandstone; 70 m³/h per well; 76 MWth installed; ~50% renewable share; replaces ~14 million m³ gas/yr (430,000 GJ); −35,000 t CO₂/yr.

The summer problem and the pilot

• Heat demand collapses in summer, so reinjection water exceeds the ideal (>50 °C), raising reinjection pumping power. An absorption chiller is justified as a reinjection-temperature-lowering device as much as a cooling device.
• Pilot: one absorption chiller at the North/A heating plant serving the SZETÁV HQ/customer-service building; nominal building cooling load 45 kW; thermal “COP-like” performance ~0.8; expected ~5 °C reduction in reinjected water temperature; auxiliary electricity from PV → near-zero-emission cooling.
• Retrofit detail: existing DH pipes to the HQ were never insulated against condensation; carrying chilled water requires re-insulation.
• Claims to be the first geothermal absorption cooling system in Hungary; 8 other Szeged circuits inject at ~50 °C, so replicable.

Economics

Geothermal heat at 16.7 EUR/GJ → cooling cost 0.075 EUR/kWh (= 16.7 × 3.6 / 1000 / 0.8) vs ~0.08 EUR/kWh for conventional AC (SCOP 3 at 0.24 EUR/kWh electricity). Near-parity, with the reinjection saving as the tie-breaker.

Literature relayed

Yilmaz (2017): NH₃-water on 100 °C / 100 kg/s geothermal in Turkey → COP 0.441, discounted payback 8.82 yr. Ketfi et al. (2023): geothermal wells as heat sink for the absorber/condenser (double-effect needs ~35% longer ground loop).

Key quotes

• “the main financial advantage of the absorption unit is that it reduces the temperature of the thermal water to be reinjected by about 5 °C. This, reduces the electricity cost of recirculation and helps maintain optimal reinjection conditions.”
• “Thus, the cost of 1 kWh of cooling energy is (16.7EUR*3,6/1000/0.8) = 0.075 EUR/kWh. Using the average SCOP = 3 of air conditioners and the current electricity price (0.24 EUR/kWh), the operating cost of conventional cooling is ~0.08 EUR/kWh.”

Companion: Geothermal Rising response to DOE RFI on AI Infrastructure on DOE Lands (7 May 2025)

Source: https://geothermal.org/sites/default/files/2025-05/Data%20Center%20RFI%20Response%20-%20Geothermal%20Rising.pdf (trade-association policy letter; qualitative, no temperatures, capacities or costs — a WebFetch summary invented figures that are not in the document)

• Proposes geothermal heat fed directly into absorption chillers for data-center cooling as the alternative to geothermal electricity → electric chillers, framed for hot, humid, water-constrained sites.
• Full cascade for DOE sites: one resource → power plant → data-center cooling → heat-intensive industry (greenhouses, aquaculture, pulp/paper, brewing, drying) → district heating; data-center low-grade waste heat fed back into the cascade.
• Quotes: “geothermal heat can be implemented directly into absorption chiller cooling systems, thereby eliminating energy waste from converting heat to electricity”; “a single geothermal heat resource can generate power at a plant, cool a data centre, and provide heat to heat-intensive industries.”
• Status: proposal only. No operating geothermal absorption-cooled data center was found; all hyperscaler geothermal deals to date (Google–Fervo, Meta–Sage, Meta–XGS) are electricity PPAs.

Relevance to the cooling wiki

Two different practitioner rationales for geothermal sorption cooling that are not “cheap cold”: in Szeged, lowering reinjection temperature to save pumping; in the Geothermal Rising letter, avoiding the heat-to-electricity conversion loss for data-center cooling. Both are real motivations; only the first has a funded pilot.