City-scale exergy-aware district heating interconnection: A GIS–graph–MCDM–MILP framework with multi-source waste heat integration

The transition to sustainable urban heating systems is a key priority in achieving decarbonization targets, particularly in densely populated cities. This study proposes a five-phase framework methodology based on the hybrid GIS, graph theory, MCDM, and optimization to integrate existing and future district heating into one large-scale district heating network through available underground corridors and by leveraging locally available renewable energy and waste heat sources specifically shallow geothermal, data centers, industrial waste heat, waste water treatment, combined heat and power, and powerplants. It will also enable an inter-system heat exchange and coalition-based operation among district heating networks. Moreover, GIS-based spatial analysis is used to identify high heat demand zones across the city and assess their proximity to heat sources and underground corridors to expand the large-scale thermal distribution along its route. The methodology evaluates candidate routing corridors and source–demand pairings based on multiple criteria, including annualized system cost, carbon emissions, exergy utilization, temperature compatibility, and spatial impedance through underground infrastructure. A TOPSIS-based ranking filters strategic corridors prior to mixed-integer linear optimization, which minimizes total annualized cost, CO₂ penalties, and exergy destruction while respecting fourth-generation DH temperature constraints. The findings for the case study in Dublin reveal that the innovative use of its existing underground infrastructure, combined with GIS-driven spatial planning, presents a transformative opportunity for scaling renewable-based district heating systems. Techno-economic and exergy optimization shows exergy efficiency rising from 28% to more than 52% for an optimized low-temperature network with heat-pump boosting, with central scenarios reducing annual operational CO₂ to 32% of the current system and optimistic designs approaching 89–94% reductions. Typical simple payback estimates fall in the range of 18 to 24 years under tunnel-reuse and heat-pump-assisted scenarios.